View in EDS HTML Full Text PDF Full Text

AI and Learning with AI: University Students' Metaphorical Conceptualizations

Saved in:
Bibliographic Details
Title: AI and Learning with AI: University Students' Metaphorical Conceptualizations
Language: English
Authors: Maria Zirenko (ORCID 0000-0003-4495-4220), Ina Alexandra Machura (ORCID 0000-0002-8192-0292), Sabine Fabriz (ORCID 0000-0003-2262-9283), Lukas Schulze-Vorberg (ORCID 0000-0003-2443-990X), Holger Horz (ORCID 0000-0002-5173-0252)
Source: Journal of Interactive Media in Education. 2025 2025(1).
Availability: Institute of Educational Technology, The Open University. Walton Hall, Milton Keynes, MK7 6AA, UK. e-mail: jime@open.ac.uk; Web site: http://jime.open.ac.uk
Peer Reviewed: Y
Page Count: 14
Publication Date: 2025
Document Type: Journal Articles
Reports - Research
Education Level: Higher Education
Postsecondary Education
Descriptors: Artificial Intelligence, Technology Uses in Education, College Students, Student Attitudes, Figurative Language, Concept Formation, Knowledge Level, Abstract Reasoning, Discourse Analysis, Computer Attitudes, Electronic Learning, Misconceptions, Foreign Countries
Geographic Terms: Germany
Abstract: The introduction of artificial intelligence (AI) in people's lives, including in educational settings, is happening rapidly and on a massive scale. However, AI represents a complicated and abstract concept for laypeople and is, in its entirety, still quite unfamiliar to many, including students in higher education. Metaphors may facilitate the comprehension of novel or abstract concepts in terms of something already known, and help investigate implicit beliefs that have the potential to influence an individual's actions. This study explored undergraduate students' (n = 124) perceptions of AI and of learning with AI by analyzing metaphors collected following an established metaphor elicitation paradigm. Students' attitudes towards AI, AI content knowledge, and usage of AI tools were assessed. The qualitative analysis of metaphors of "AI" yielded nine categories (e.g., "brain, human, machinery, unknown"), while the analysis of metaphors for "learning with AI" yielded seven categories (e.g., "self-regulation, educator, shared learning"). Overall, the anthropomorphization of AI for both foci was observed. Many conceptualized "learning with AI" as learning with trustworthy support, and foregrounded the perceived facilitation of learning on the basis of AI. This study highlights the importance of fostering accurate conceptualizations of AI and its role in learning, while addressing misconceptions and overly simplistic representations. Promoting a nuanced understanding of AI is essential to ensuring its effective use as a tool that enhances, rather than impedes, learning processes.
Abstractor: As Provided
Entry Date: 2025
Accession Number: EJ1483180
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwFTqUNpDuakh0EcZNH2zn6MAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDEtKZZ8qaf9j5ELTdQIBEICBmzQFydtjku8Spk5j_EvjeVOOaMvVQ9A0p44am0YiNpvGFtpqoyyheMz2wFzn3Q1bQXu-B7RKARBZ4EjvE9mpB_Mpp-WzfCXUt0nPto8gswp4MNbQpRQ1tXQlY3YnYbtufC4qvf8xvWDrjGilJ7hnnzd5wGIm32PH-j6W9QdZAe8XA6sjhNTu94DM43-dcCZLGRfkD1UMJ9cz4mWu
Text:
  Availability: 1
  Value: <anid>AN0190586076;[1ctx]01jan.25;2026Jan02.02:28;v2.2.500</anid> <title id="AN0190586076-1">AI and Learning with AI: University Students' Metaphorical Conceptualizations </title> <p>The introduction of artificial intelligence (AI) in people's lives, including in educational settings, is happening rapidly and on a massive scale. However, AI represents a complicated and abstract concept for laypeople and is, in its entirety, still quite unfamiliar to many, including students in higher education. Metaphors may facilitate the comprehension of novel or abstract concepts in terms of something already known, and help investigate implicit beliefs that have the potential to influence an individual's actions. This study explored undergraduate students' (n = 124) perceptions of AI and of learning with AI by analyzing metaphors collected following an established metaphor elicitation paradigm. Students' attitudes towards AI, AI content knowledge, and usage of AI tools were assessed. The qualitative analysis of metaphors of AI yielded nine categories (e.g., brain, human, machinery, unknown), while the analysis of metaphors for learning with AI yielded seven categories (e.g., self-regulation, educator, shared learning). Overall, the anthropomorphization of AI for both foci was observed. Many conceptualized learning with AI as learning with trustworthy support, and foregrounded the perceived facilitation of learning on the basis of AI. This study highlights the importance of fostering accurate conceptualizations of AI and its role in learning, while addressing misconceptions and overly simplistic representations. Promoting a nuanced understanding of AI is essential to ensuring its effective use as a tool that enhances, rather than impedes, learning processes.</p> <p>Keywords: artificial intelligence; AI in Education; metaphors for AI; metaphors for learning with AI; higher education</p> <hd id="AN0190586076-2">Introduction</hd> <p>Artificial intelligence (AI) remains a relatively novel and abstract phenomenon for many higher education students. Yet, it is simultaneously becoming a tool that university learners increasingly employ for various purposes. For instance, the Digital Education Council Global AI Student Survey 2024, conducted among students across 16 countries and covering academic levels from undergraduate to doctoral studies in various disciplines, indicates that 86% of students are already utilizing AI applications such as ChatGPT, Google Gemini, or Microsoft Copilot in their academic studies ([<reflink idref="bib18" id="ref1">18</reflink>]), while 69% of students reported using AI primarily for information searches. Inevitably, individuals differ in their knowledge of what AI is, how AI-based applications work, and their attitude towards AI. AI literacy assessments reveal a need for more in-depth knowledge among university students as documented by performance-based evaluations ([<reflink idref="bib28" id="ref2">28</reflink>]) and self-report measures ([<reflink idref="bib40" id="ref3">40</reflink>]), highlighting a disparity between AI competence and AI usage, with students in technical disciplines outperforming those in the social sciences and humanities ([<reflink idref="bib40" id="ref4">40</reflink>]). Since the widespread use of AI among students remains a relatively recent development, it can be assumed that many learners have not yet developed formal knowledge of AI, and rather rely on a superficial or naïve understanding of its design and functionalities. As AI becomes increasingly integrated into students' learning experiences, primarily through their independent use of generative AI tools, it is crucial to investigate how learners perceive and conceptualize AI within educational settings. Learner conceptualizations for AI may shape how learners use AI, and influence how it is employed within educational environments. While a large body of research focuses on students' perceptions of and attitudes towards AI in education, spanning topics from general AI to specific applications ([<reflink idref="bib16" id="ref5">16</reflink>]), metaphor studies are less frequently employed to articulate abstract conceptualizations of AI, even though metaphors for AI are common in both verbal and visual discourses ([<reflink idref="bib10" id="ref6">10</reflink>]; [<reflink idref="bib29" id="ref7">29</reflink>]), including discourses on education.</p> <hd id="AN0190586076-3">AI in Education (AIEd)</hd> <p>Artificial Intelligence in Education (AIEd) is not a new concept, but in recent years – especially with the advent of generative AI tools – it has quickly gained prominence as a substantial area of research and innovation within educational technology ([<reflink idref="bib11" id="ref8">11</reflink>]). Research shows that students generally have a positive attitude toward integrating generative AI into their educational experiences ([<reflink idref="bib14" id="ref9">14</reflink>]). Additionally, university students with greater knowledge and experience with current technology in general tend to exhibit more positive attitudes toward AI applications such as ChatGPT ([<reflink idref="bib25" id="ref10">25</reflink>]). However, studies show that naïve understandings of AI 'encompass' misconceptions. In their review of 25 studies covering primary to tertiary education, Bewersdorff et al. ([<reflink idref="bib7" id="ref11">7</reflink>]) found that learners often lack both technical knowledge and a realistic understanding of AI's capabilities and limitations. The authors observed a tendency for learners to ascribe human-like attributes to AI, and noted that their perceptions of its benefits and risks were often binary and lacking in nuance. To combat misleading assumptions about AI and to promote a proper understanding of its use, educators and practitioners have positioned <emph>AI literacy</emph> as a necessary competency for the future ([<reflink idref="bib22" id="ref12">22</reflink>]).</p> <hd id="AN0190586076-4">Metaphors</hd> <p>Metaphors have the potential to shape language and cognition by drawing analogies between different concepts ([<reflink idref="bib24" id="ref13">24</reflink>]; [<reflink idref="bib37" id="ref14">37</reflink>]). In this sense, metaphors are understood not only as stylistic figures in creative language use but, in the case of <emph>conceptual</emph> metaphors, as linguistic expressions of <emph>cognitive</emph> analogies in which a source domain (e.g., <emph>a journey</emph>) helps to understand a target domain (e.g., <emph>a relationship</emph>, as in 'Look how far we've come together'; see, e.g., [<reflink idref="bib38" id="ref15">38</reflink>]). Thus, metaphors allow us to relate unfamiliar, complex, abstract ideas to familiar experiences, e.g., computational systems to brains ([<reflink idref="bib12" id="ref16">12</reflink>]). As metaphors posit correspondences between two different domains, they highlight certain characteristics of these domains while obscuring others ([<reflink idref="bib23" id="ref17">23</reflink>]; [<reflink idref="bib37" id="ref18">37</reflink>]).</p> <p>A range of studies show that metaphors regulate thoughts and actions as well as frame communication of messages. This can affect how people generate solutions to complex problems based on conceptualizations consistent with the metaphorical frame and, consequently, nudge action-taking or decision-making in a specific direction ([<reflink idref="bib24" id="ref19">24</reflink>]; [<reflink idref="bib57" id="ref20">57</reflink>]). Other studies do not find the same evidence (e.g., [<reflink idref="bib53" id="ref21">53</reflink>]) or suggest differential effects of metaphor acceptance ([<reflink idref="bib44" id="ref22">44</reflink>]).</p> <hd id="AN0190586076-5">Metaphors for AI in education</hd> <p>Metaphors have also been used as proxies for tacit knowledge to investigate implicit aspects of cognition, especially in the field of education, where metaphors are assumed to be related to belief systems about learning ([<reflink idref="bib1" id="ref23">1</reflink>]; [<reflink idref="bib58" id="ref24">58</reflink>]) or teaching ([<reflink idref="bib52" id="ref25">52</reflink>]). They have been empirically examined to gain insight into the belief systems of teachers ([<reflink idref="bib8" id="ref26">8</reflink>]) and learners ([<reflink idref="bib39" id="ref27">39</reflink>]; [<reflink idref="bib47" id="ref28">47</reflink>]; [<reflink idref="bib59" id="ref29">59</reflink>]). For instance, Sfard ([<reflink idref="bib50" id="ref30">50</reflink>]) proposed two 'lenses' on higher education teaching and learning: the <emph>acquisition metaphor of learning</emph> and the <emph>participation metaphor</emph> as an enculturation into a subject community. Similarly, Wegner and Nückles ([<reflink idref="bib60" id="ref31">60</reflink>]) differentiated between metaphors foregrounding <emph>regulation, acquisition, problem-solving</emph>, or <emph>personality development</emph> within their student data. Importantly, groups of students who provided different types of metaphors differed in motivation and use of deep processing strategies ([<reflink idref="bib60" id="ref32">60</reflink>]). For educational technologies, metaphors have been used historically to guide both technology design and implementation, e.g., as "building <emph>blocks</emph> or empty <emph>vessels</emph>" ([<reflink idref="bib23" id="ref33">23</reflink>] italics in the original). Stevenson ([<reflink idref="bib55" id="ref34">55</reflink>]) revealed <emph>resource, tutor, tool</emph>, and <emph>environment</emph> as categories for metaphors of digital technology in pedagogy. For AI usage in education, Lim ([<reflink idref="bib39" id="ref35">39</reflink>]) explored metaphors among pre-service teachers and categorized the data into positive (e.g., <emph>assistant teacher</emph>) and negative metaphorical concepts (e.g., <emph>double-sided meaning</emph>).</p> <p>Further, metaphors have been utilized to qualitatively assess mental models of AI as a novel concept for many ([<reflink idref="bib19" id="ref36">19</reflink>]; [<reflink idref="bib41" id="ref37">41</reflink>]). For instance, if students compare AI to human entities, such as <emph>teachers</emph> or <emph>assistants</emph>, they might notice specific functionalities and limitations of AI ([<reflink idref="bib3" id="ref38">3</reflink>]). Existing, yet still sparse, research investigates metaphors of AI in different populations using metaphor elicitation procedures. In a sample of Chinese undergraduates, Xu and colleagues identified <emph>uncontrollable unknown, auxiliary helpers</emph>, and <emph>advanced algorithms</emph> as common metaphors ([<reflink idref="bib61" id="ref39">61</reflink>]). Another study with a Turkish student sample found conceptualizations as, e.g., <emph>human, assistance</emph>, or <emph>growing sprout</emph> ([<reflink idref="bib3" id="ref40">3</reflink>]). In a similar vein, Kavak and Yılmaz ([<reflink idref="bib31" id="ref41">31</reflink>]) grouped metaphors elicited among library science professionals into five themes (e.g., a relational theme with <emph>friend</emph> or <emph>child</emph>, or a functional theme with <emph>servant</emph> or <emph>help desk</emph>).</p> <p>How individuals conceptualize AI is closely linked to their attitudes toward it ([<reflink idref="bib9" id="ref42">9</reflink>]). Research has shown that presented metaphors may shape users' expectations and behaviors toward AI systems even before direct interaction. For instance, users to whom an AI agent is introduced as a <emph>toddler</emph> or a <emph>professional</emph> can exhibit substantially different expectations regarding its competence and reliability ([<reflink idref="bib32" id="ref43">32</reflink>]). This fits with evidence on how the choice of visual metaphors for conversational agents can arouse uncertainty or trust toward the agents ([<reflink idref="bib30" id="ref44">30</reflink>]). Depending on whether metaphors align with users' expectations, they can either encourage or deter users from engaging with AI.</p> <hd id="AN0190586076-6">Research questions</hd> <p>As noted earlier, the substantial integration of AI into educational contexts is a relatively recent and rapidly evolving development for learners. Consequently, it is essential to explore its impact on how individual learners conceptualize and engage with AI. In this regard, metaphors offer a valuable tool for understanding perceptions of AI. This study aims to examine students' perceptions of AI and their experiences of learning with AI. It employs a classification of metaphorical expressions to explore how learners perceive AI and how they tend to use it.</p> <p> <emph>RQ 1: Which metaphors do university students use to describe the concept of 'AI' and the concept of 'learning with AI'?</emph> </p> <p>In addition, the present study seeks to document potential relationships between the participants' metaphorical imagery and their attitudes, knowledge, and usage of AI.</p> <p> <emph>RQ 2: Are attitudes to AI, AI content knowledge, and AI usage related to the metaphors participants use to describe AI and learning with AI?</emph> </p> <hd id="AN0190586076-7">Method</hd> <p></p> <hd id="AN0190586076-8">Sample</hd> <p>A total of 136 students at a German university participated in the study (84.1% female, 11.9% male, 2.4% non-binary/diverse, mean age = 22.88 years, <emph>SD</emph> = 4.21). Most were bachelor's students (89.7%). Four students reported a background in AI (content in studies, work, knowledge in R). Twelve participants were excluded from the analysis due to missing data, resulting in <emph>n</emph> = 124. The study was administered as an online survey in German using SoSciSurvey. The link was disseminated through psychology and teacher education courses. Psychology students could earn credit for their participation. Data was collected from 28 February to 12 July 2024. The study did not require individual ethical review but received ethical approval from the local ethics committee based on its general characteristics.</p> <hd id="AN0190586076-9">Measures and analysis</hd> <p></p> <hd id="AN0190586076-10">Elicitation, identification, and categorization of metaphors</hd> <p>For the elicitation of metaphors, an established paradigm was followed ([<reflink idref="bib47" id="ref45">47</reflink>]; [<reflink idref="bib60" id="ref46">60</reflink>]). Participants were asked to complete a sentence with a metaphor and provide a brief explanation for their formulation. To familiarize participants with this procedure, an example of a metaphor unrelated to the research topic was provided: <emph>"Language is like a bridge because it connects people."</emph> Participants were then prompted to think of a metaphor representing AI (<emph>"AI is like ..., because ..."</emph>) and learning with AI (<emph>"Learning with AI is like ..., because ..."</emph>). To enrich responses for the latter, the additional prompt, <emph>"The goal of learning with AI is ...,"</emph> was included, also following the procedure of Wegner and Nückles ([<reflink idref="bib60" id="ref47">60</reflink>]). To <emph>identify</emph> non-metaphorical responses, the <emph>metaphor identification procedure</emph> (MIP) by the Pragglejaz Group ([<reflink idref="bib27" id="ref48">27</reflink>]) was implemented.</p> <p>For metaphor <emph>categorization</emph>, a range of existing coding schemes was reviewed ([<reflink idref="bib3" id="ref49">3</reflink>]; [<reflink idref="bib19" id="ref50">19</reflink>]; [<reflink idref="bib39" id="ref51">39</reflink>]; [<reflink idref="bib61" id="ref52">61</reflink>]). However, none of these schemes were adopted in the present study due to concerns regarding the potential cultural dependence of metaphors, the limited generalizability of the coding frameworks, or lack of information on intercoder reliability. Consequently, a data-driven semantic approach was employed to develop a new coding scheme ([<reflink idref="bib42" id="ref53">42</reflink>]; [<reflink idref="bib60" id="ref54">60</reflink>]).</p> <hd id="AN0190586076-11">Metaphors for AI</hd> <p>For <emph>metaphor identification</emph>, two authors initially applied the MIP to all responses independently, resolving any discrepancies through discussion until consensus was reached. The analysis revealed that not all responses instantiated metaphors as intended with the elicitation task. As a result, an additional coding category, <emph>different metaphor</emph>, was introduced for cases where participants conceptualized personal experiences with AI rather than AI itself (<emph>n</emph> = 12, e.g., 'a journey through time into the future'). Also, a category for <emph>not a metaphor</emph> was added for cases where the responses did not qualify as metaphorical expressions (<emph>n</emph> = 16, e.g., 'internet knowledge processing'). Responses categorized as either <emph>different metaphor</emph> or <emph>not a metaphor</emph>, as well as missing responses (<emph>n</emph> = 5) were excluded from further analysis, leaving only valid AI metaphors for consideration (<emph>n</emph> = 91).</p> <p>For <emph>metaphor categorization</emph>, two authors independently applied a data-driven semantic approach to group metaphorical expressions in accordance with their semantic similarity ([<reflink idref="bib42" id="ref55">42</reflink>]; [<reflink idref="bib60" id="ref56">60</reflink>]). Disagreements were resolved through discussion. A third author independently applied the finalized coding scheme (see Appendix 1). Interrater reliability was assessed using Fleiss' κ, yielding a value of 0.813.</p> <hd id="AN0190586076-12">Metaphors for learning with AI</hd> <p>Following the procedure outlined above, a preliminary analysis was conducted to identify responses that did not qualify as metaphors but instead provided literal explanations of learning with AI (<emph>n</emph> = 38, e.g., 'learning with help'), as well as metaphors unrelated to learning with AI (<emph>different metaphors, n</emph> = 7, e.g., 'a subconscious process'). These 45 responses were excluded from further analysis, resulting in 79 valid metaphors for learning with AI. Initially, an existing coding scheme for <emph>metaphors of learning</emph> ([<reflink idref="bib60" id="ref57">60</reflink>]) was considered for the present data set. However, during the coding process, it became apparent that this scheme would require substantial modification to capture the broader array of meanings in the data set, as metaphors for <emph>learning with AI</emph> involved themes that extended beyond those associated with <emph>learning</emph> as such. As a result, the data-driven development of categories was informed by, but not confined to, the four categories proposed by Wegner and Nückles ([<reflink idref="bib60" id="ref58">60</reflink>]). Disagreements were discussed and resolved through consensus. Subsequently, all responses were independently coded by a third author. Responses that did not fit the eight established categories were coded as <emph>miscellaneous</emph>. Interrater reliability was substantial, with a Fleiss' κ of 0.60.</p> <hd id="AN0190586076-13">Attitudes towards AI</hd> <p>Attitudes were measured using the <emph>Attitude towards Artificial Intelligence</emph> (ATTARI-12) scale ([<reflink idref="bib54" id="ref59">54</reflink>]). The scale comprises cognitive, affective, and behavioral facets measured as one construct. The 12 items are rated on a five-point scale (1 = <emph>strongly disagree</emph> to 5 = <emph>strongly agree</emph>). The instructions were adapted by excluding the definition of AI because we were interested in measuring attitudes toward the conceptual understanding of AI that students currently possess. Cronbach's ɑ was 0.90.</p> <hd id="AN0190586076-14">AI content knowledge</hd> <p>To counter the effects of self-reported knowledge ([<reflink idref="bib4" id="ref60">4</reflink>]), AI content knowledge (AI-CK) was evaluated using a single open-ended item: <emph>In the following, we are interested in what you understand by the term 'artificial intelligence'. Please describe in a few sentences what you understand by it</emph>. Answers should reflect participants' basic understanding of AI as an aspect of their AI-CK ([<reflink idref="bib33" id="ref61">33</reflink>]). The responses were each divided into coherent segments (<emph>k</emph> = 169), and each segment was categorized according to a categorization scheme with 11 categories (e.g., <emph>machine learning</emph> or <emph>example of an AI application</emph>; quantitative content analysis, [<reflink idref="bib35" id="ref62">35</reflink>]). This scheme was initially developed deductively by one of the authors and revised inductively based on the empirical responses. Two independent raters evaluated the responses in each segment for correctness with good interrater reliability (Cohen's κ = 0.883). The responses were quantified, awarding one point for each correct segment.</p> <hd id="AN0190586076-15">Usage of AI tools</hd> <p>AI tool usage was assessed using a composite index that included four components: (a) self-reported frequency of AI tool use in three different contexts – personal, academic, and vocational – each rated on a six-point scale (1 = <emph>not at all</emph> to 6 = <emph>very frequent</emph>), and (b) the total number of AI tools used, based on a provided list with the option to add additional tools. These four indicators were combined to create the overall index.</p> <hd id="AN0190586076-16">Results</hd> <p></p> <hd id="AN0190586076-17">RQ 1: Analysis of metaphors</hd> <p></p> <hd id="AN0190586076-18">Metaphors for AI</hd> <p>The qualitative analysis of the metaphors for AI yielded nine categories; metaphors that did not match these categories and could not form another category were labeled <emph>miscellaneous</emph> (see Table 1; see Appendix 1 for a full description of the categories). To explore relations with the quantitative variables, the nine categories were summarized based on semantic similarities into four overarching <emph>broad categories</emph>.</p> <p></p> <ulist> <item> <emph>Functionality</emph> refers to how AI is used or functionally described. Includes the categories <emph>utility</emph> (AI as a tool) and <emph>machinery</emph> (AI as a machine or mechanical tool).</item> <p></p> <item> <emph>Knowledge and intelligence</emph> refers to AI's computational potential and data handling. Includes the categories <emph>brain</emph> (AI as a thinking machine or brain) and <emph>knowledge</emph> (AI as a knowledge repository or encyclopedia).</item> <p></p> <item> <emph>Human similarity</emph> represents AI in a human role. Includes categories <emph>human</emph> (AI as a human or human-like actor), <emph>authority</emph> (AI as a normative or authoritative agent), <emph>interaction</emph> (AI as a process that enables interactions).</item> <p></p> <item> <emph>Ambivalence and uncertainty</emph> highlights the unknown or ambiguous aspects of AI. Includes <emph>double-sided meaning</emph> (positive and negative potentials of AI in one metaphor) and <emph>unknown</emph> (AI as something unknown or difficult to predict).</item> </ulist> <p>Table 1 Categories for metaphors of AI with examples and percentages.</p> <p> <ephtml> <table><tr><th colspan="5" /></tr><tr><th align="left" valign="top">BROAD CATEGORY</th><th align="left" valign="top">CATEGORY</th><th align="left" valign="top">EXAMPLES</th><th align="left" valign="top">PERCENTAGE OF CATEGORY WITHIN ALL METAPHORS FOR AI</th><th align="left" valign="top">PERCENTAGE OF BROAD CATEGORY WITHIN METAPHORS FOR AI INCLUDED IN BROAD CATEGORIES</th></tr><tr><th colspan="5" /></tr><tr><td align="left" valign="top" rowspan="3"><p>Functionality</p></td><td align="left" valign="top"><p>Utility</p></td><td align="left" valign="top"><p>a toolbox</p></td><td align="left" valign="top"><p>13.2%</p></td><td align="left" valign="top"><p>25.6%</p></td></tr><tr><td colspan="4" /></tr><tr><td align="left" valign="top"><p>Machinery</p></td><td align="left" valign="top"><p>a crane, a drone</p></td><td align="left" valign="top"><p>9.9%</p></td><td align="left" valign="top" /></tr><tr><td colspan="5" /></tr><tr><td align="left" valign="top" rowspan="3"><p>Knowledge and intelligence</p></td><td align="left" valign="top"><p>Brain</p></td><td align="left" valign="top"><p>perfected brain, a brain with no capacity limit, the brain of a machine</p></td><td align="left" valign="top"><p>14.3%</p></td><td align="left" valign="top"><p>30.5%</p></td></tr><tr><td colspan="4" /></tr><tr><td align="left" valign="top"><p>Knowledge</p></td><td align="left" valign="top"><p>an encyclopedia, a searchable lexicon</p></td><td align="left" valign="top"><p>13.2%</p></td><td align="left" valign="top" /></tr><tr><td colspan="5" /></tr><tr><td align="left" valign="top" rowspan="5"><p>Human similarity</p></td><td align="left" valign="top"><p>Human</p></td><td align="left" valign="top"><p>a toddler, a good colleague</p></td><td align="left" valign="top"><p>14.3%</p></td><td align="left" valign="top"><p>23.2%</p></td></tr><tr><td colspan="4" /></tr><tr><td align="left" valign="top"><p>Authority</p></td><td align="left" valign="top"><p>an omniscient authority, a possibly uncontrollable superpower</p></td><td align="left" valign="top"><p>4.4%</p></td><td align="left" valign="top" /></tr><tr><td colspan="4" /></tr><tr><td align="left" valign="top"><p>Interaction</p></td><td align="left" valign="top"><p>a study group</p></td><td align="left" valign="top"><p>2.2%</p></td><td align="left" valign="top" /></tr><tr><td colspan="5" /></tr><tr><td align="left" valign="top" rowspan="3"><p>Ambivalence and uncertainty</p></td><td align="left" valign="top"><p>Double-sided meaning</p></td><td align="left" valign="top"><p>a double-edged sword, Icarus' wings</p></td><td align="left" valign="top"><p>8.8%</p></td><td align="left" valign="top"><p>20.7%</p></td></tr><tr><td colspan="4" /></tr><tr><td align="left" valign="top"><p>Unknown</p></td><td align="left" valign="top"><p>an interstellar probe in the vastness of space, a forbidden door</p></td><td align="left" valign="top"><p>9.9%</p></td><td align="left" valign="top" /></tr><tr><td colspan="5" /></tr><tr><td align="left" valign="top"><p>–</p></td><td align="left" valign="top"><p>Miscellaneous</p></td><td align="left" valign="top"><p>an ant colony, a sponge</p></td><td align="left" valign="top"><p>9.9%</p></td><td align="left" valign="top"><p>–</p></td></tr><tr><td colspan="5" /></tr></table> </ephtml> </p> <hd id="AN0190586076-19">Metaphors for learning with AI</hd> <p>Based on the complete answer (i.e., metaphor and explanation), seven categories of metaphors were extracted from the data, with an added <emph>miscellaneous</emph> category for remaining cases (see Table 2; see Appendix 2 for a full description of the categories).</p> <p>Table 2 Categories for metaphors of learning with AI with examples and percentages.</p> <p> <ephtml> <table><tr><th colspan="5" /></tr><tr><th align="left" valign="top">BROAD CATEGORY</th><th align="left" valign="top">CATEGORY</th><th align="left" valign="top">EXAMPLES</th><th align="left" valign="top">PERCENTAGE OF CATEGORY WITHIN ALL METAPHORS FOR LEARNING WITH AI</th><th align="left" valign="top">PERCENTAGE OF BROAD CATEGORY WITHIN ALL METAPHORS FOR LEARNING WITH AI INCLUDED IN BROAD CATEGORIES</th></tr><tr><th colspan="5" /></tr><tr><td align="left" valign="top"><p>AI as a tool for self- improvement</p></td><td align="left" valign="top"><p>Self-regulation</p></td><td align="left" valign="top"><p>cycling with training wheels</p></td><td align="left" valign="top"><p>29.1%</p></td><td align="left" valign="top"><p>47.8%</p></td></tr><tr><td colspan="5" /></tr><tr><td align="left" valign="top" /><td align="left" valign="top"><p>Knowledge acquisition</p></td><td align="left" valign="top"><p>learning with an interactive textbook</p></td><td align="left" valign="top"><p>6.3%</p></td><td align="left" valign="top" /></tr><tr><td colspan="5" /></tr><tr><td align="left" valign="top" /><td align="left" valign="top"><p>Extension</p></td><td align="left" valign="top"><p>learning with a second brain</p></td><td align="left" valign="top"><p>5.1%</p></td><td align="left" valign="top" /></tr><tr><td colspan="5" /></tr><tr><td align="left" valign="top" rowspan="5"><p>AI as an interaction with a partner or supporter</p></td><td align="left" valign="top"><p>Assistance</p></td><td align="left" valign="top"><p>live like a king, I have my own personal helpers who give me advice and knowledge</p></td><td align="left" valign="top"><p>10.1%</p></td><td align="left" valign="top"><p>52.2%</p></td></tr><tr><td colspan="4" /></tr><tr><td align="left" valign="top"><p>Shared learning</p></td><td align="left" valign="top"><p>learning with a friend</p></td><td align="left" valign="top"><p>8.9%</p></td><td align="left" valign="top" /></tr><tr><td colspan="4" /></tr><tr><td align="left" valign="top"><p>Educator/instructor</p></td><td align="left" valign="top"><p>teacher</p></td><td align="left" valign="top"><p>26.6%</p></td><td align="left" valign="top" /></tr><tr><td colspan="5" /></tr><tr><td align="left" valign="top"><p>–</p></td><td align="left" valign="top"><p>Double-sided meaning</p></td><td align="left" valign="top"><p>train travel with German Railways: sometimes easy, sometimes unreliable</p></td><td align="left" valign="top"><p>2.5%</p></td><td align="left" valign="top"><p>–</p></td></tr><tr><td colspan="5" /></tr><tr><td align="left" valign="top"><p>–</p></td><td align="left" valign="top"><p>Miscellaneous</p></td><td align="left" valign="top"><p>the search for the needle in the haystack</p></td><td align="left" valign="top"><p>11.4%</p></td><td align="left" valign="top"><p>–</p></td></tr><tr><td colspan="5" /></tr></table> </ephtml> </p> <p>The categories were then grouped into two <emph>broad categories</emph> in accordance with semantic similarity. The <emph>double-sided meaning</emph> category could not be classified into either broad category, and therefore was excluded from the subsequent analysis.</p> <p></p> <ulist> <item> <emph>AI as a tool (for self-improvement)</emph> presents AI as a tool to enhance or improve learning processes. Includes the categories <emph>self-regulation, knowledge acquisition</emph>, and <emph>extension</emph>.</item> <p></p> <item> <emph>AI as an interaction with a partner or supporter</emph> casts AI as taking on an active role, assisting or cooperating with the learner. Includes the categories <emph>assistance, shared learning</emph>, and <emph>educator/instructor</emph>.</item> </ulist> <hd id="AN0190586076-20">Relationships between metaphors for AI and for learning with AI</hd> <p>We performed a Chi-Square test to explore whether the broad metaphor categories for AI and learning with AI were related to one another and found a significant result (χ2 (<reflink idref="bib3" id="ref63">3</reflink>) = 12.401, <emph>p</emph> < 0.001).</p> <p>Participants whose metaphorical conceptions fall into the realms of <emph>functionality</emph> or <emph>human similarity</emph> represented learning with AI as an 'interaction with a partner'.</p> <p>Participants whose metaphors for AI correspond to <emph>ambivalence and uncertainty</emph> thought of learning with AI as a <emph>tool for self-improvement</emph> and not as <emph>interaction with a partner</emph>. Participants who produced AI metaphors that can be grouped as <emph>knowledge and intelligence</emph> did not show preference for any type of metaphor for learning with AI.</p> <hd id="AN0190586076-21">RQ2: Differential analysis of metaphors</hd> <p></p> <hd id="AN0190586076-22">Descriptive statistics for qualitative variables and correlation analysis</hd> <p>To check for the interrelatedness of the quantitative measures, a correlation analysis was performed (see Table 3, also for descriptives). A positive moderate correlation (<emph>r</emph> = 0.46, <emph>p</emph> < 0.01) was observed between the usage of AI tools and attitude toward AI, such that a more positive attitude relates to a higher usage value. AI content knowledge was not significantly correlated to either attitude or AI usage. Importantly, the measure for AI content knowledge showed limited variance in the study sample and may, therefore, be interpreted only with caution.</p> <p>Table 3 Descriptive statistics and correlations among quantitative measures.</p> <p> <ephtml> <table><tr><th colspan="7" /></tr><tr><th align="left" valign="top" /><th align="left" valign="top"><italic>n</italic></th><th align="left" valign="top"><italic>M</italic></th><th align="left" valign="top"><italic>SD</italic></th><th align="left" valign="top">1</th><th align="left" valign="top">2</th><th align="left" valign="top">3</th></tr><tr><th colspan="7" /></tr><tr><td align="left" valign="top"><p>1. Attitudes towards AI</p></td><td align="left" valign="top"><p>123</p></td><td align="left" valign="top"><p>3.48</p></td><td align="left" valign="top"><p>0.63</p></td><td align="left" valign="top"><p>–</p></td><td align="left" valign="top" /><td align="left" valign="top" /></tr><tr><td colspan="7" /></tr><tr><td align="left" valign="top"><p>2. AI-CK</p></td><td align="left" valign="top"><p>124</p></td><td align="left" valign="top"><p>1.08</p></td><td align="left" valign="top"><p>0.88</p></td><td align="left" valign="top"><p>0.720</p></td><td align="left" valign="top"><p>–</p></td><td align="left" valign="top" /></tr><tr><td colspan="7" /></tr><tr><td align="left" valign="top"><p>3. Usage of AI tools (index)</p></td><td align="left" valign="top"><p>123</p></td><td align="left" valign="top"><p>12.39</p></td><td align="left" valign="top"><p>3.79</p></td><td align="left" valign="top"><p>0.464**</p></td><td align="left" valign="top"><p>0.970</p></td><td align="left" valign="top"><p>–</p></td></tr><tr><td colspan="7" /></tr></table> </ephtml> </p> <p>1 Note. **<emph>p</emph> < 0.01.</p> <p>Subsequently, potential relationships between AI knowledge, attitudes, usage, and broad metaphor category were investigated.</p> <hd id="AN0190586076-23">Metaphors for AI</hd> <p>An ANOVA did not reveal significant differences between the broad categories for AI metaphors in terms of attitude towards AI, <emph>F</emph>(<reflink idref="bib3" id="ref64">3</reflink>, 78) = 1.016; <emph>p</emph> = 0.390, AI knowledge, <emph>F</emph>(<reflink idref="bib3" id="ref65">3</reflink>, 78) = 2.188; <emph>p</emph> = 0.096, or AI usage, <emph>F</emph>(<reflink idref="bib3" id="ref66">3</reflink>, 78) = 1.091; <emph>p</emph> = 0.358.</p> <hd id="AN0190586076-24">Metaphors for learning with AI</hd> <p>An ANOVA for the broad categories of learning with AI metaphors did not reveal significant differences based on category for either attitude towards AI, <emph>F</emph>(<reflink idref="bib1" id="ref67">1</reflink>, 64) = 1.162; <emph>p</emph> = 0.285, AI knowledge, <emph>F</emph>(<reflink idref="bib1" id="ref68">1</reflink>, 64) = 2.568; <emph>p</emph> = 0.114, or AI usage, <emph>F</emph>(<reflink idref="bib1" id="ref69">1</reflink>,<reflink idref="bib64" id="ref70">64</reflink>) = 0.001; <emph>p</emph> = 0.974.</p> <hd id="AN0190586076-25">Discussion</hd> <p>This study analyzed metaphors to investigate university students' conceptualizations of AI and of learning with AI in higher education. The findings offer insights into the students' understandings of AI design and functions, the role of AI in learning processes, and into the relationships between metaphorical mappings, knowledge of, and attitude toward AI.</p> <hd id="AN0190586076-26">Metaphors for AI</hd> <p>Our first research question addressed the categories of metaphors students use to describe AI. First, it was found that students use anthropomorphizing imagery to conceptualize AI, which is in line with observations among information professionals ([<reflink idref="bib31" id="ref71">31</reflink>]), middle school students ([<reflink idref="bib3" id="ref72">3</reflink>]; [<reflink idref="bib19" id="ref73">19</reflink>]), and university students in Asia ([<reflink idref="bib61" id="ref74">61</reflink>]; [<reflink idref="bib63" id="ref75">63</reflink>]). The <emph>brain/neural functioning</emph> metaphors observed in the present data set were also common among these groups. Thinking of AI as "an alien person rather than a human-built machine" ([<reflink idref="bib43" id="ref76">43</reflink>]) can make its use more intuitive and appealing. This perspective could also be deliberately employed by developers to enhance user engagement and confidence ([<reflink idref="bib48" id="ref77">48</reflink>]). By the same token, as <emph>brain/neural</emph> imagery is frequent in popular discourse and science communication about AI ([<reflink idref="bib10" id="ref78">10</reflink>]; [<reflink idref="bib46" id="ref79">46</reflink>]), a more substantial similarity between the architecture and functioning of the human brain and AI applications may be suggested than is the case, which in turn may impact user behavior. Complementarily, metaphors that highlight specifically human-like attributes of AI, such as potential autonomy, have already been discussed as obscuring the substantial imprint of actual humans in the design, deployment, and functioning of AI applications ([<reflink idref="bib29" id="ref80">29</reflink>]).</p> <p> <emph>Tool & utility</emph> metaphors also appear to emerge from a tradition of public discourse that uses physical tools (e.g., hammers) to conceptualize technological advances, such as AI, and previous advances, such as personal computers or the internet. Interestingly, these metaphors shape not only the discourse of laypeople but are also adopted by technological experts when explaining technological advances to the public ([<reflink idref="bib21" id="ref81">21</reflink>]).</p> <p>Metaphors of <emph>authority</emph> that may potentially inspire problematic levels of trust among users were documented not only on our student data set but also, interestingly, among information professionals ([<reflink idref="bib31" id="ref82">31</reflink>]).</p> <p>Metaphors that conceptualize AI as similar to <emph>knowledge and information repositories</emph> were frequent in the present as well as in other studies: "AI is like a book with an unlimited number of pages because you can get any information you need" (present study), a "digital library" ([<reflink idref="bib31" id="ref83">31</reflink>]). These conceptions of AI as a technology where information is stored and can be accessed in the same form recast metaphorical language laypersons use to conceptualize the internet ([<reflink idref="bib62" id="ref84">62</reflink>]). Thus, learners using AI might be tempted to transfer understandings erroneously from one digital technology to another. Importantly, studies have demonstrated that students in schools and universities generally overestimate their ability to fact-check digital information ([<reflink idref="bib45" id="ref85">45</reflink>]), which in turn might result in consequences when students conceptualize AI applications as knowledge repositories akin to the internet.</p> <p>There is a noteworthy lack of the <emph>generative</emph> aspect of AI observed in the present study. The generative element is altogether rare in metaphor elicitation studies concerning AI. Existing examples in elicited data sets are few ("AI is like a painter because each data point creates a combination of different colors and patterns on a canvas", elicited among information professionals; [<reflink idref="bib31" id="ref86">31</reflink>]), and are observed among novices mostly in response to designs that specifically mention the generative aspect ([<reflink idref="bib63" id="ref87">63</reflink>]). In the present study, none of the elicited metaphors evoked a generative source domain. This lack of generative imagery might indicate that, so far, the generative aspect of AI is not reflected in thinking about AI. Moreover, previous research has demonstrated the challenges laypeople face in utilizing understandings of statistical probabilities for their reasoning, e.g., when evaluating digital information ([<reflink idref="bib2" id="ref88">2</reflink>]). Thus, metaphorical conceptions that hint at a lack of knowledge about AI's generative, probabilistic architecture underscore the need for supporting AI literacy in students. Students who are already using AI in their studies while knowing comparatively little about it might mistake the seemingly effortless interaction with AI engines as a case of ecological rationality – a situation in which students feel they understand AI well enough to integrate it into their learning processes ([<reflink idref="bib17" id="ref89">17</reflink>]).</p> <p>Four broad categories of metaphors for AI identified in the present study, namely <emph>functionality, knowledge and intelligence, human similarity</emph>, and <emph>ambivalence and uncertainty</emph>, mostly overlap with those published elsewhere ([<reflink idref="bib3" id="ref90">3</reflink>]; [<reflink idref="bib31" id="ref91">31</reflink>]; [<reflink idref="bib61" id="ref92">61</reflink>]). This suggests that beliefs about AI are shared rather than widely different across contexts.</p> <hd id="AN0190586076-27">Metaphors for learning with AI</hd> <p>Among metaphors for learning with AI, seven categories (excluding a <emph>miscellaneous</emph> category) and two broad categories were identified. The role of AI in learning was explored by comparing the metaphor categories identified with those commonly found in similar contexts. For example, Wegner and Nückles' ([<reflink idref="bib60" id="ref93">60</reflink>]) four metaphors of learning were partially reflected in the present data. Metaphors related to <emph>regulation</emph> and <emph>knowledge acquisition</emph> were observed in the present data set, but not those associated with learning as <emph>problem-solving</emph> or <emph>personality development</emph>. These latter metaphors may emerge as students gain more experience with learning with AI. Our findings also align with previous studies on metaphors for digital technology in education, which describe it as a tool, resource, and tutor ([<reflink idref="bib55" id="ref94">55</reflink>]).</p> <p>Within the <emph>self-regulation</emph> category, metaphors reflecting support, self-improvement, and reliance on AI for achieving learning goals were observed in the present study. This understanding is consistent with the assumption that AI can facilitate self-regulation in ways previously unattainable, a hypothesis that is the subject of extensive current research ([<reflink idref="bib5" id="ref95">5</reflink>]). Another well-represented category of <emph>educator/instructor</emph> might also indicate reliance on AI support in learning. However, evidence suggests that individuals may over-rely on AI-generated advice, even when it contradicts contextual information or their own judgment ([<reflink idref="bib34" id="ref96">34</reflink>]). Contrary to digital literacy discourses promoting fact-checking AI responses, people align their opinions with fact-checks flagged as AI-generated ([<reflink idref="bib13" id="ref97">13</reflink>]).</p> <p>Broad categorizations of metaphors for learning with AI revealed two main conceptual groups: <emph>AI as a tool</emph> and <emph>AI as a partner or supporter</emph>. These categories differ from one another in the degree of agency and autonomy attributed to AI. Additionally, analyzing responses that were not coded as valid metaphors revealed further insights. Several participants expressed concerns about threats to academic integrity, diminishing critical thinking skills, and the risk of overreliance on AI. These concerns were also reported among participants in previous research ([<reflink idref="bib63" id="ref98">63</reflink>]). Thus, participants in the current study were aware that AI may be used for academic cheating. This behavior is linked to Dark Triad personality traits ([<reflink idref="bib26" id="ref99">26</reflink>]).</p> <hd id="AN0190586076-28">Usage and knowledge of AI</hd> <p>Among the current participants, AI usage and AI knowledge did not correlate. Notably, the concept of 'digital natives' has long been criticized for falsely assuming that frequent use of digital technology inevitably fosters ICT literacy, including AI literacy. Research has shown that this assumption is oversimplified, as factors such as access to digital technology, the extent of its use, or its use in school only account for a small portion of the variance in ICT skills ([<reflink idref="bib49" id="ref100">49</reflink>]).</p> <p>In contrast to studies that observe quantitative relationships between metaphorical categories, e.g., for learning on the one hand, and measures of attitude and/or behavior on the other ([<reflink idref="bib58" id="ref101">58</reflink>]), no such relationships emerged in the present study. Possibly, some participants might have produced metaphors that do not necessarily represent their own conceptualizations of AI, but instead imagery the participants had encountered in public or academic discourse about AI and learning with AI (e.g., 'a neural network').</p> <hd id="AN0190586076-29">Implications for research and practice</hd> <p>This study provides insights into the diverse categories of metaphors students use to conceptualize AI and learning with AI. Additionally, it shows that students' metaphorical conceptions are not necessarily connected to their level of knowledge about AI or to their usage. Also, research indicates that presenting metaphorical explanations for digital technology does not necessarily lead to correct conceptual shifts ([<reflink idref="bib36" id="ref102">36</reflink>]). Accordingly, more research is needed to inform educational strategies based on metaphorical material.</p> <p>At the same time, building AI literacy is a key strategy for reducing misconceptions among learners ([<reflink idref="bib33" id="ref103">33</reflink>]; [<reflink idref="bib51" id="ref104">51</reflink>]), specifically as those misconceptions of AI persist even with substantial usage and are found at all levels of the educational system.</p> <hd id="AN0190586076-30">Limitations</hd> <p>The current study has several limitations. The sample consisted predominantly of psychology undergraduates. The findings may vary in more disciplinarily diverse groups or other populations, as well as across generations since acceptance and use of generative AI differ between different age groups ([<reflink idref="bib15" id="ref105">15</reflink>]). Future research could build on the present insights by contrasting metaphors elicited among different populations (higher education faculty, K-12 educators, or students) and taking into account the participants' diverse cultural and linguistic profiles. The stability of the results over time should also be carefully considered, particularly in light of the rapid advancements in AI and growing literacy concerning AI-powered technologies.</p> <p>Besides, difficulties in eliciting actual metaphors from participants are commonly reported in the empirical literature on conceptual metaphors. 12% of responses to the 'AI' prompt and 32% to the 'learning with AI' prompt were coded as <emph>not a metaphor</emph> in the present study. Occasionally, researchers select rates as low as 30% of participants' responses to metaphor elicitation prompts (see [<reflink idref="bib64" id="ref106">64</reflink>]) for further analysis. It is possible that some responses in the present study could not be categorized as actual metaphors due to the rapidity of AI development and its relatively recent integration into education: students may still need to develop their concepts of AI in learning environments in order to be able to articulate their conceptions in metaphorical expressions. Another possible explanation could be individual differences that influence the production of metaphors (e.g., cognitive abilities; [<reflink idref="bib6" id="ref107">6</reflink>]). The observed lack of the generative aspect in the metaphors could also be connected to the chosen method of metaphor elicitation. Further studies could explore other methods such as metaphor construction tasks or image drawing (e.g., [<reflink idref="bib52" id="ref108">52</reflink>]).</p> <p>Depending on whether participants have used generative chatbots or other applications, their conceptualizations and metaphorical imagery may differ, and future research can address this.</p> <p>The methodology in the present study adapts traditional procedures of metaphor identification (MIP, [<reflink idref="bib27" id="ref109">27</reflink>]). Another option could be to adopt AI-based procedures for training different Large Language Models (LLMs), such as BERT, to identify metaphors in written discourse ([<reflink idref="bib56" id="ref110">56</reflink>]). It is also unclear how to transfer existing applications to different languages.</p> <p>In future studies, the metaphor categorization scheme based on sematic proximity could be combined with, e.g., valence ratings. Consistent valence ratings cannot necessarily be expected in studies where novel metaphors abound, and a substantial variety of metaphors are produced. By comparison, procedures to train LLMs to assess metaphor creativity are piloted, but still need refinement ([<reflink idref="bib20" id="ref111">20</reflink>]).</p> <hd id="AN0190586076-31">Conclusion</hd> <p>The present study analyzed the categories of metaphors used by university students to conceptualize AI and learning with AI. Even as the metaphors reveal a lack of awareness concerning the generative element of AI applications, no meaningful associations emerged between metaphorical conceptualizations and either AI knowledge or usage. Consistent with previous research, however, the findings demonstrate that frequent use of AI does not necessarily go hand in hand with a deeper understanding of the technology. These results suggest how the assumption that 'learning by doing' alone is sufficient for developing competent AI usage may be misguided. This study underscores the critical need to promote realistic conceptualizations of AI in educational contexts as AI increasingly influences teaching and learning processes. Fostering AI literacy is essential to encouraging the use of AI in ways that enhance, rather than hinder, student learning. Misunderstandings or oversimplified views of AI can constrain its educational potential, making it essential to address these through targeted educational interventions.</p> <p>While metaphors may offer insights into how learners perceive and engage with AI technologies, their elicitation and analysis are not without methodological and practical challenges. Future research will need to validate the assumptions underlying research designs that rely on metaphor elicitation.</p> <p>This research advocates for a broader and more nuanced approach to studying the psychology of AI. Future efforts should take into account the diverse ways in which learners understand AI, the methods used to assess these understandings, and the cultural contexts in which learning takes place. By considering these factors, educators and researchers can more effectively leverage AI's potential to foster meaningful and equitable learning experiences.</p> <hd id="AN0190586076-32">Data Accessibility Statement</hd> <p>Please email the corresponding author to obtain raw data.</p> <hd id="AN0190586076-33">Additional File</hd> <p>The additional file for this article can be found as follows:</p> <p>Graph: Supplementary Information Appendix 1 and 2. DOI:</p> <hd id="AN0190586076-34">Competing Interests</hd> <p>The authors have no competing interests to declare.</p> <hd id="AN0190586076-35">Author contributions</hd> <p>Maria Zirenko, Ina Alexandra Machura, Sabine Fabriz, and Lukas Schulze-Vorberg contributed to the design and implementation of the research, to the analysis of the results and to the writing of the manuscript. Holger Horz contributed to the final version of the manuscript and supervised the project. All authors read and approved the final manuscript.</p> <ref id="AN0190586076-36"> <title> REFERENCES </title> <blist> <bibl id="bib1" idref="ref23" type="bt">1</bibl> <bibtext> Adano, C and Bunn, G. 2024. 'Taking the plunge, juggling acts, and friendly fire: Metaphors that distance learning students use to describe their experiences of online learning'. Journal of Perspectives in Applied Academic Practice, 12(1). DOI: 10.56433/jpaap.v12i1.566</bibtext> </blist> <blist> <bibl id="bib2" idref="ref88" type="bt">2</bibl> <bibtext> Álvarez-Arroyo, R, Batanero, C and Gea, MM. 2024. 'Probabilistic literacy and reasoning of prospective secondary school teachers when interpreting media news'. ZDM – Mathematics Education, 56(6): 1045–1058. DOI: 10.1007/s11858-024-01586-8</bibtext> </blist> <blist> <bibl id="bib3" idref="ref38" type="bt">3</bibl> <bibtext> Aydin, S, Duman, E, Bertiz, Y and Birişçi, S. 2022. 'Investigation of the effects of computer-aided animations on conceptual understanding through metaphors: An example of artificial intelligence'. Journal of Educational Technology and Online Learning, 5(4): 1140–1159. DOI: 10.31681/jetol.1151460</bibtext> </blist> <blist> <bibl id="bib4" idref="ref60" type="bt">4</bibl> <bibtext> Baier-Mosch, F, Schulze-Vorberg, L, Horz, H and Kunter, M. 2024. 'Exploring the suitability of self-reports to gain insights into teachers' technological pedagogical knowledge'. Studies in Educational Evaluation, 83: 101402. DOI: 10.1016/j.stueduc.2024.101402</bibtext> </blist> <blist> <bibl id="bib5" idref="ref95" type="bt">5</bibl> <bibtext> Banihashem, SK, Bond, M, Bergdahl, N, Khosravi, H and Noroozi, O. 2024. A systematic mapping review at the intersection of artificial intelligence and self-regulated learning: A call for increased theoretical grounding, focus on motivation and diversifying context. [Preprint] DOI: 10.13140/RG.2.2.13057.47207</bibtext> </blist> <blist> <bibl id="bib6" idref="ref107" type="bt">6</bibl> <bibtext> Beaty, RE and Silvia, PJ. 2013. 'Metaphorically speaking: Cognitive abilities and the production of figurative language'. Memory & Cognition, 41(2): 255–267. DOI: 10.3758/s13421-012-0258-5</bibtext> </blist> <blist> <bibl id="bib7" idref="ref11" type="bt">7</bibl> <bibtext> Bewersdorff, A, Zhai, X, Roberts, J and Nerdel, C. 2023. 'Myths, mis- and preconceptions of artificial intelligence: A review of the literature'. Computers and Education: Artificial Intelligence, 4: 100143. DOI: 10.1016/j.caeai.2023.100143</bibtext> </blist> <blist> <bibl id="bib8" idref="ref26" type="bt">8</bibl> <bibtext> Blau, I, Grinberg, R and Shamir-Inbal, T. 2018. 'Pedagogical perspectives and practices reflected in metaphors of learning and digital learning of ICT leaders'. Computers in the Schools, 35(1): 32–48. DOI: 10.1080/07380569.2018.1427960</bibtext> </blist> <blist> <bibl id="bib9" idref="ref42" type="bt">9</bibl> <bibtext> Bochniarz, KT, Czerwiński, SK, Sawicki, A and Atroszko, PA. 2022. 'Attitudes to AI among high school students: Understanding distrust towards humans will not help us understand distrust towards AI'. Personality and Individual Differences, 185: 111299. DOI: 10.1016/j.paid.2021.111299</bibtext> </blist> <blist> <bibtext> Boucher, P. 2021. What if we chose new metaphors for artificial intelligence. Bruxelles: EPRSE-European Parliamentary Research Service.</bibtext> </blist> <blist> <bibtext> Bozkurt, A, Xiao, J, Farrow, R, Bai, JYH, Nerantzi, C, Moore, S, Dron, J, Stracke, CM, Singh, L, Crompton, H, Koutropoulos, A, Terentev, E, Pazurek, A, Nichols, M, Sidorkin, AM, Costello, E, Watson, S, Mulligan, D and Honeychurch, S, et al. 2024. 'The manifesto for teaching and learning in a time of generative AI: A critical collective stance to better navigate the future'. Open Praxis, 16(4): 487–513. DOI: 10.55982/openpraxis.16.4.777</bibtext> </blist> <blist> <bibtext> Carbonell, J, Sánchez-Esguevillas, A, and Carro, B. (2016). 'The role of metaphors in the development of technologies. The case of the artificial intelligence'. Futures, 84: 145–153. DOI: 10.1016/j.futures.2016.03.019</bibtext> </blist> <blist> <bibtext> Chae, JH and Tewksbury, D. 2024. Perceiving AI intervention does not compromise the persuasive effect of fact-checking. New Media & Society, Article 14614448241286881. DOI: 10.1177/14614448241286881</bibtext> </blist> <blist> <bibtext> Chan, CKY and Hu, W. 2023. 'Students' voices on generative AI: Perceptions, benefits, and challenges in higher education'. International Journal of Educational Technology in Higher Education, 20(43). DOI: 10.1186/s41239-023-00411-8</bibtext> </blist> <blist> <bibtext> Chan, CKY and Lee, KKW. 2023. 'The AI generation gap: Are Gen Z students more interested in adopting generative AI such as ChatGPT in teaching and learning than their Gen X and millennial generation teachers?'Smart Learning Environments, 10(60). DOI: 10.1186/s40561-023-00269-3</bibtext> </blist> <blist> <bibtext> Chan, CKY and Zhou, W. 2023. 'An expectancy value theory (EVT) based instrument for measuring student perceptions of generative AI'. Smart Learning Environments, 10(64). DOI: 10.1186/s40561-023-00284-4</bibtext> </blist> <blist> <bibtext> Davidson, S. 2024. 'The economic institutions of artificial intelligence'. Journal of Institutional Economics, 20. DOI: 10.1017/S1744137423000395</bibtext> </blist> <blist> <bibtext> DEC. 2024. Digital Education Council Global AI Student Survey 2024. Digital Education Council [Online]. Available at: https://<ulink href="http://www.digitaleducationcouncil.com/post/digital-education-council-global-ai-student-survey-2024">www.digitaleducationcouncil.com/post/digital-education-council-global-ai-student-survey-2024</ulink>.</bibtext> </blist> <blist> <bibtext> Demir, K and Güraksin, GE. 2022. 'Determining middle school students' perceptions of the concept of artificial intelligence: A metaphor analysis'. Participatory Educational Research, 9(2): 297–312. DOI: 10.17275/per.22.41.9.2</bibtext> </blist> <blist> <bibtext> DiStefano, PV, Patterson, JD and Beaty, RE. 2024. 'Automatic scoring of metaphor creativity with large language models'. Creativity Research Journal, 1–15. DOI: 10.1080/10400419.2024.2326343</bibtext> </blist> <blist> <bibtext> Droog, E, Burgers, C and Kee, KF. 2020. 'How journalists and experts metaphorically frame emerging information technologies: The case of cyberinfrastructure for big data'. Public Understanding of Science, 29(8): 819–834. DOI: 10.1177/0963662520952542</bibtext> </blist> <blist> <bibtext> European Commission (EU). 2024. EU artificial intelligence act; Article 4: AI literacy [Online]. Available at: https://artificialintelligenceact.eu/article/4/.</bibtext> </blist> <blist> <bibtext> Ferreira, GMD, Lemgruber, MS and Cabrera, TL. 2023. 'From didachography to AI: Metaphors teaching is automated by'. Journal of Interactive Media in Education, 2023(1). DOI: 10.5334/jime.798</bibtext> </blist> <blist> <bibtext> Flusberg, SJ, Holmes, KJ, Thibodeau, PH, Nabi, RL and Matlock, T. 2024. 'The psychology of framing: How everyday language shapes the way we think, feel, and act'. Psychological Science in the Public Interest: A Journal of the American Psychological Society, 25(3): 105–161. DOI: 10.1177/15291006241246966</bibtext> </blist> <blist> <bibtext> García-Alonso, EM, León-Mejía, AC, Sánchez-Cabrero, R and Guzmán-Ordaz, R. 2024. 'Training and technology acceptance of ChatGPT in university students of social sciences: A netcoincidental analysis'. Behavioral Sciences, 14(7). DOI: 10.3390/bs14070612</bibtext> </blist> <blist> <bibtext> Greitemeyer, T and Kastenmüller, A. 2023. 'HEXACO, the Dark Triad, and Chat GPT: Who is willing to commit academic cheating?'Heliyon, 9(9): e19909. DOI: 10.1016/j.heliyon.2023.e19909</bibtext> </blist> <blist> <bibtext> Group, P. 2007. 'MIP: A method for identifying metaphorically used words in discourse'. Metaphor and Symbol, 22(1): 1–39. DOI: 10.1080/10926480709336752</bibtext> </blist> <blist> <bibtext> Hornberger, M, Bewersdorff, A and Nerdel, C. 2023. 'What do university students know about artificial intelligence? Development and validation of an AI literacy test'. Computers and Education: Artificial Intelligence, 5: 100165. DOI: 10.1016/j.caeai.2023.100165</bibtext> </blist> <blist> <bibtext> Johnson, DG and Verdicchio, M. 2017. 'Reframing AI discourse'. Minds and Machines, 27(4): 575–590. DOI: 10.1007/s11023-017-9417-6</bibtext> </blist> <blist> <bibtext> Jung, J-Y, Qiu, S, Bozzon, A and Gadiraju, U. 2022. 'Great chain of agents: The role of metaphorical representation of agents in conversational crowdsourcing text'. In: CHI Conference on Human Factors in Computing Systems, New York, ACM. pp. 1–22. DOI: 10.1145/3491102.3517653</bibtext> </blist> <blist> <bibtext> Kavak, A and Yılmaz, E. 2024. 'Information professionals' metaphorical perceptions of artificial intelligence concept'. Journal of Library Administration, 64(6): 695–718. DOI: 10.1080/01930826.2024.2371276</bibtext> </blist> <blist> <bibtext> Khadpe, P, Krishna, R, Fei-Fei, L, Hancock, JT and Bernstein, MS. 2020. 'Conceptual metaphors impact perceptions of human-AI collaboration'. Proceedings of the ACM on Human-Computer Interaction, 4(CSCW2): 1–26. DOI: 10.1145/3415234</bibtext> </blist> <blist> <bibtext> Kim, K, Kwon, K, Ottenbreit-Leftwich, A, Bae, H and Glazewski, K. 2023. 'Exploring middle school students' common naive conceptions of artificial intelligence concepts, and the evolution of these ideas'. Education and Information Technologies, 28(8): 9827–9854. DOI: 10.1007/s10639-023-11600-3</bibtext> </blist> <blist> <bibtext> Klingbeil, A, Grützner, C and Schreck, P. 2024. 'Trust and reliance on AI — An experimental study on the extent and costs of overreliance on AI'. Computers in Human Behavior, 160: 108352. DOI: 10.1016/j.chb.2024.108352</bibtext> </blist> <blist> <bibtext> Krippendorf, K. 1989. 'Content analysis'. In: Barnouw, E, Gerbner, G, Schramm, W, Worth, TL and Gross, L (eds.) International encyclopedia of communication [Online]. New York: Oxford University Press, pp. 403–407. Available at: <ulink href="http://repository.upenn.edu/asc%5fpapers/226">http://repository.upenn.edu/asc%5fpapers/226</ulink>.</bibtext> </blist> <blist> <bibtext> Kulyk, O, Renaud, K and Costica, S. 2023. 'People want reassurance when making privacy-related decisions — Not technicalities'. Journal of Systems and Software, 200: 111620. DOI: 10.1016/j.jss.2023.111620</bibtext> </blist> <blist> <bibtext> Lakoff, G and Johnson, M. 2008. Metaphors we live by [with a new afterword], 5th ed. Chicago, Ill: The University of Chicago Press.</bibtext> </blist> <blist> <bibtext> Landau, MJ and Keefer, LA. 2014. 'This is like that: Metaphors in public discourse shape attitudes'. Social and Personality Psychology Compass, 8(8): 463–473. DOI: 10.1111/spc3.12125</bibtext> </blist> <blist> <bibtext> Lim, EM. 2024. 'Metaphor analysis on pre-service early childhood teachers' conception of AI (Artificial Intelligence) education for young children'. Thinking Skills and Creativity, 51: 101455. DOI: 10.1016/j.tsc.2023.101455</bibtext> </blist> <blist> <bibtext> Mansoor, HMH, Bawazir, A, Alsabri, MA, Alharbi, A and Okela, AH. 2024. 'Artificial intelligence literacy among university students—a comparative transnational survey'. Frontiers in Communication, 9. DOI: 10.3389/fcomm.2024.1478476</bibtext> </blist> <blist> <bibtext> Marx, E, Leonhardt, T and Bergner, N. 2023. 'Secondary school students' mental models and attitudes regarding artificial intelligence – a scoping review'. Computers and Education: Artificial Intelligence, 5: 100169. DOI: 10.1016/j.caeai.2023.100169</bibtext> </blist> <blist> <bibtext> Mayring, P. 2015. 'Qualitative content analysis: Theoretical background and procedures'. In: Bikner-Ahsbahs, A, Knipping, C and Presmeg, N (eds.) Advances in mathematics education. Approaches to qualitative research in mathematics education. Dordrecht: Springer, pp. 365–380. DOI: 10.1007/978-94-017-9181-6_13</bibtext> </blist> <blist> <bibtext> Mollick, E. 2024. Co-intelligence: Living and working with AI. Penguin.</bibtext> </blist> <blist> <bibtext> Panzeri, F, Di Paola, S and Domaneschi, F. 2021. 'Does the COVID-19 war metaphor influence reasoning?'PloS one, 16(4): e0250651. DOI: 10.1371/journal.pone.0250651</bibtext> </blist> <blist> <bibtext> Porat, E, Blau, I and Barak, A. 2018. 'Measuring digital literacies: Junior high-school students' perceived competencies versus actual performance'. Computers & Education, 126: 23–36. DOI: 10.1016/j.compedu.2018.06.030</bibtext> </blist> <blist> <bibtext> Richards, BA and Lillicrap, TP. 2022. 'The brain-computer metaphor debate is useless: A matter of semantics'. Frontiers in Computer Science, 4. DOI: 10.3389/fcomp.2022.810358</bibtext> </blist> <blist> <bibtext> Saban, A, Kocbeker, BN and Saban, A. 2007. 'Prospective teachers' conceptions of teaching and learning revealed through metaphor analysis'. Learning and Instruction, 17(2): 123–139. DOI: 10.1016/j.learninstruc.2007.01.003</bibtext> </blist> <blist> <bibtext> Salles, A, Evers, K and Farisco, M. 2020. 'Anthropomorphism in AI'. AJOB Neuroscience, 11(2): 88–95. DOI: 10.1080/21507740.2020.1740350</bibtext> </blist> <blist> <bibtext> Senkbeil, M. (2022). 'ICT-related variables as predictors of ICT literacy beyond intelligence and prior achievement'. Education and Information Technologies, 27(3): 3595–3622. DOI: 10.1007/s10639-021-10759-x</bibtext> </blist> <blist> <bibtext> Sfard, A. 1998. 'On two metaphors for learning and the dangers of choosing just one'. Educational Researcher, 27(2): 4–13. DOI: 10.3102/0013189X027002004</bibtext> </blist> <blist> <bibtext> Song, Y, Weisberg, LR, Zhang, S, Tian, X, Boyer, KE and Israel, M. 2024. 'A framework for inclusive AI learning design for diverse learners'. Computers and Education: Artificial Intelligence, 6: 100212. DOI: 10.1016/j.caeai.2024.100212</bibtext> </blist> <blist> <bibtext> Soysal, Y. 2024. 'Drawing a portrayal of science teachers' epistemic cognitions around different concepts characterizing science education'. Science & Education. DOI: 10.1007/s11191-023-00494-9</bibtext> </blist> <blist> <bibtext> Steen, GJ, Reijnierse, WG and Burgers, C. 2014. 'When do natural language metaphors influence reasoning? A follow-up study to Thibodeau and Boroditsky (2013)'. PloS one, 9(12): e113536. DOI: 10.1371/journal.pone.0113536</bibtext> </blist> <blist> <bibtext> Stein, J-P, Messingschlager, T, Gnambs, T, Hutmacher, F and Appel, M. 2024. 'Attitudes towards AI: Measurement and associations with personality'. Scientific Reports, 14(1): 2909. DOI: 10.1038/s41598-024-53335-2</bibtext> </blist> <blist> <bibtext> Stevenson, I. 2008. 'Tool, tutor, environment or resource: Exploring metaphors for digital technology and pedagogy using activity theory'. Computers & Education, 51(2): 836–853. DOI: 10.1016/j.compedu.2007.09.001</bibtext> </blist> <blist> <bibtext> Su, C, Fukumoto, F, Huang, X, Li, J, Wang, R and Chen, Z. 2020. 'DeepMet: A reading comprehension paradigm for token-level metaphor detection'. In: Klebanov, BB, Shutova, E, Lichtenstein, P, Muresan, S, Wee, C, Feldman, A and Ghosh, D (eds.) Proceedings of the Second Workshop on Figurative Language Processing. Stroudsburg, PA, USA: Association for Computational Linguistics. pp. 30–39. DOI: 10.18653/v1/2020.figlang-1.4</bibtext> </blist> <blist> <bibtext> Thibodeau, PH and Boroditsky, L. 2011. 'Metaphors we think with: The role of metaphor in reasoning'. PloS one, 6(2): e16782. DOI: 10.1371/journal.pone.0016782</bibtext> </blist> <blist> <bibtext> Wegner, E, Burkhart, C and Nückles, M. 2021. 'The short-term reliability of metaphors as an assessment method'. Frontiers in Education, 6: 678327. DOI: 10.3389/feduc.2021.678327</bibtext> </blist> <blist> <bibtext> Wegner, E, Burkhart, C, Weinhuber, M and Nückles, M. 2020. 'What metaphors of learning can (and cannot) tell us about students' learning'. Learning and Individual Differences, 80: 101884. DOI: 10.1016/j.lindif.2020.101884</bibtext> </blist> <blist> <bibtext> Wegner, E and Nückles, M. 2015. 'From eating to discovering: How metaphors of learning change during students' enculturation'. Zeitschrift für Hochschulentwicklung, 10(4). DOI: 10.3217/zfhe-10-04/08</bibtext> </blist> <blist> <bibtext> Xu, L, Zhang, Y, Yu, F, Ding, X and Wu, J. 2024. 'Folk beliefs of artificial intelligence and robots'. International Journal of Social Robotics, 16(3): 429–446. DOI: 10.1007/s12369-024-01097-2</bibtext> </blist> <blist> <bibtext> Yaman, İ. 2016. 'Exploring prospective English language teachers' perceptions of the internet through metaphorical conceptualizations'. Educational Research and Reviews, 11(8): 499–507. DOI: 10.5897/ERR2016.2681</bibtext> </blist> <blist> <bibtext> Yan, Y, Sun, W and Zhao, X. 2024. 'Metaphorical conceptualizations of generative artificial intelligence use by Chinese university EFL learners'. Frontiers in Education, 9. DOI: 10.3389/feduc.2024.1430494</bibtext> </blist> <blist> <bibtext> Zhang, X. 2021. 'Development and critiques of conceptual metaphor theory'. Theory and Practice in Language Studies, 11(11): 1487–1491. DOI: 10.17507/tpls.1111.18</bibtext> </blist> </ref> <aug> <p>By Maria Zirenko; Ina Alexandra Machura; Sabine Fabriz; Lukas Schulze-Vorberg and Holger Horz</p> <p>Reported by Author; Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib18" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib28" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib40" firstref="ref3"></nolink> <nolink nlid="nl4" bibid="bib16" firstref="ref5"></nolink> <nolink nlid="nl5" bibid="bib10" firstref="ref6"></nolink> <nolink nlid="nl6" bibid="bib29" firstref="ref7"></nolink> <nolink nlid="nl7" bibid="bib11" firstref="ref8"></nolink> <nolink nlid="nl8" bibid="bib14" firstref="ref9"></nolink> <nolink nlid="nl9" bibid="bib25" firstref="ref10"></nolink> <nolink nlid="nl10" bibid="bib22" firstref="ref12"></nolink> <nolink nlid="nl11" bibid="bib24" firstref="ref13"></nolink> <nolink nlid="nl12" bibid="bib37" firstref="ref14"></nolink> <nolink nlid="nl13" bibid="bib38" firstref="ref15"></nolink> <nolink nlid="nl14" bibid="bib12" firstref="ref16"></nolink> <nolink nlid="nl15" bibid="bib23" firstref="ref17"></nolink> <nolink nlid="nl16" bibid="bib57" firstref="ref20"></nolink> <nolink nlid="nl17" bibid="bib53" firstref="ref21"></nolink> <nolink nlid="nl18" bibid="bib44" firstref="ref22"></nolink> <nolink nlid="nl19" bibid="bib58" firstref="ref24"></nolink> <nolink nlid="nl20" bibid="bib52" firstref="ref25"></nolink> <nolink nlid="nl21" bibid="bib39" firstref="ref27"></nolink> <nolink nlid="nl22" bibid="bib47" firstref="ref28"></nolink> <nolink nlid="nl23" bibid="bib59" firstref="ref29"></nolink> <nolink nlid="nl24" bibid="bib50" firstref="ref30"></nolink> <nolink nlid="nl25" bibid="bib60" firstref="ref31"></nolink> <nolink nlid="nl26" bibid="bib55" firstref="ref34"></nolink> <nolink nlid="nl27" bibid="bib19" firstref="ref36"></nolink> <nolink nlid="nl28" bibid="bib41" firstref="ref37"></nolink> <nolink nlid="nl29" bibid="bib61" firstref="ref39"></nolink> <nolink nlid="nl30" bibid="bib31" firstref="ref41"></nolink> <nolink nlid="nl31" bibid="bib32" firstref="ref43"></nolink> <nolink nlid="nl32" bibid="bib30" firstref="ref44"></nolink> <nolink nlid="nl33" bibid="bib27" firstref="ref48"></nolink> <nolink nlid="nl34" bibid="bib42" firstref="ref53"></nolink> <nolink nlid="nl35" bibid="bib54" firstref="ref59"></nolink> <nolink nlid="nl36" bibid="bib33" firstref="ref61"></nolink> <nolink nlid="nl37" bibid="bib35" firstref="ref62"></nolink> <nolink nlid="nl38" bibid="bib64" firstref="ref70"></nolink> <nolink nlid="nl39" bibid="bib63" firstref="ref75"></nolink> <nolink nlid="nl40" bibid="bib43" firstref="ref76"></nolink> <nolink nlid="nl41" bibid="bib48" firstref="ref77"></nolink> <nolink nlid="nl42" bibid="bib46" firstref="ref79"></nolink> <nolink nlid="nl43" bibid="bib21" firstref="ref81"></nolink> <nolink nlid="nl44" bibid="bib62" firstref="ref84"></nolink> <nolink nlid="nl45" bibid="bib45" firstref="ref85"></nolink> <nolink nlid="nl46" bibid="bib17" firstref="ref89"></nolink> <nolink nlid="nl47" bibid="bib34" firstref="ref96"></nolink> <nolink nlid="nl48" bibid="bib13" firstref="ref97"></nolink> <nolink nlid="nl49" bibid="bib26" firstref="ref99"></nolink> <nolink nlid="nl50" bibid="bib49" firstref="ref100"></nolink> <nolink nlid="nl51" bibid="bib36" firstref="ref102"></nolink> <nolink nlid="nl52" bibid="bib51" firstref="ref104"></nolink> <nolink nlid="nl53" bibid="bib15" firstref="ref105"></nolink> <nolink nlid="nl54" bibid="bib56" firstref="ref110"></nolink> <nolink nlid="nl55" bibid="bib20" firstref="ref111"></nolink>
CustomLinks:
  – Url: https://eric.ed.gov/contentdelivery/servlet/ERICServlet?accno=EJ1483180
    Name: ERIC Full Text
    Category: fullText
    Text: Full Text from ERIC
Header DbId: eric
DbLabel: ERIC
An: EJ1483180
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: AI and Learning with AI: University Students' Metaphorical Conceptualizations
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Maria+Zirenko%22">Maria Zirenko</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0003-4495-4220">0000-0003-4495-4220</externalLink>)<br /><searchLink fieldCode="AR" term="%22Ina+Alexandra+Machura%22">Ina Alexandra Machura</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-8192-0292">0000-0002-8192-0292</externalLink>)<br /><searchLink fieldCode="AR" term="%22Sabine+Fabriz%22">Sabine Fabriz</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0003-2262-9283">0000-0003-2262-9283</externalLink>)<br /><searchLink fieldCode="AR" term="%22Lukas+Schulze-Vorberg%22">Lukas Schulze-Vorberg</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0003-2443-990X">0000-0003-2443-990X</externalLink>)<br /><searchLink fieldCode="AR" term="%22Holger+Horz%22">Holger Horz</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-5173-0252">0000-0002-5173-0252</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Journal+of+Interactive+Media+in+Education%22"><i>Journal of Interactive Media in Education</i></searchLink>. 2025 2025(1).
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Institute of Educational Technology, The Open University. Walton Hall, Milton Keynes, MK7 6AA, UK. e-mail: jime@open.ac.uk; Web site: http://jime.open.ac.uk
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 14
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Higher+Education%22">Higher Education</searchLink><br /><searchLink fieldCode="EL" term="%22Postsecondary+Education%22">Postsecondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Artificial+Intelligence%22">Artificial Intelligence</searchLink><br /><searchLink fieldCode="DE" term="%22Technology+Uses+in+Education%22">Technology Uses in Education</searchLink><br /><searchLink fieldCode="DE" term="%22College+Students%22">College Students</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Attitudes%22">Student Attitudes</searchLink><br /><searchLink fieldCode="DE" term="%22Figurative+Language%22">Figurative Language</searchLink><br /><searchLink fieldCode="DE" term="%22Concept+Formation%22">Concept Formation</searchLink><br /><searchLink fieldCode="DE" term="%22Knowledge+Level%22">Knowledge Level</searchLink><br /><searchLink fieldCode="DE" term="%22Abstract+Reasoning%22">Abstract Reasoning</searchLink><br /><searchLink fieldCode="DE" term="%22Discourse+Analysis%22">Discourse Analysis</searchLink><br /><searchLink fieldCode="DE" term="%22Computer+Attitudes%22">Computer Attitudes</searchLink><br /><searchLink fieldCode="DE" term="%22Electronic+Learning%22">Electronic Learning</searchLink><br /><searchLink fieldCode="DE" term="%22Misconceptions%22">Misconceptions</searchLink><br /><searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink>
– Name: Subject
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Germany%22">Germany</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: The introduction of artificial intelligence (AI) in people's lives, including in educational settings, is happening rapidly and on a massive scale. However, AI represents a complicated and abstract concept for laypeople and is, in its entirety, still quite unfamiliar to many, including students in higher education. Metaphors may facilitate the comprehension of novel or abstract concepts in terms of something already known, and help investigate implicit beliefs that have the potential to influence an individual's actions. This study explored undergraduate students' (n = 124) perceptions of AI and of learning with AI by analyzing metaphors collected following an established metaphor elicitation paradigm. Students' attitudes towards AI, AI content knowledge, and usage of AI tools were assessed. The qualitative analysis of metaphors of "AI" yielded nine categories (e.g., "brain, human, machinery, unknown"), while the analysis of metaphors for "learning with AI" yielded seven categories (e.g., "self-regulation, educator, shared learning"). Overall, the anthropomorphization of AI for both foci was observed. Many conceptualized "learning with AI" as learning with trustworthy support, and foregrounded the perceived facilitation of learning on the basis of AI. This study highlights the importance of fostering accurate conceptualizations of AI and its role in learning, while addressing misconceptions and overly simplistic representations. Promoting a nuanced understanding of AI is essential to ensuring its effective use as a tool that enhances, rather than impedes, learning processes.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2025
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1483180
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1483180
RecordInfo BibRecord:
  BibEntity:
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 14
    Subjects:
      – SubjectFull: Artificial Intelligence
        Type: general
      – SubjectFull: Technology Uses in Education
        Type: general
      – SubjectFull: College Students
        Type: general
      – SubjectFull: Student Attitudes
        Type: general
      – SubjectFull: Figurative Language
        Type: general
      – SubjectFull: Concept Formation
        Type: general
      – SubjectFull: Knowledge Level
        Type: general
      – SubjectFull: Abstract Reasoning
        Type: general
      – SubjectFull: Discourse Analysis
        Type: general
      – SubjectFull: Computer Attitudes
        Type: general
      – SubjectFull: Electronic Learning
        Type: general
      – SubjectFull: Misconceptions
        Type: general
      – SubjectFull: Foreign Countries
        Type: general
      – SubjectFull: Germany
        Type: general
    Titles:
      – TitleFull: AI and Learning with AI: University Students' Metaphorical Conceptualizations
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Maria Zirenko
      – PersonEntity:
          Name:
            NameFull: Ina Alexandra Machura
      – PersonEntity:
          Name:
            NameFull: Sabine Fabriz
      – PersonEntity:
          Name:
            NameFull: Lukas Schulze-Vorberg
      – PersonEntity:
          Name:
            NameFull: Holger Horz
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2025
          Numbering:
            – Type: volume
              Value: 2025
            – Type: issue
              Value: 1
          Titles:
            – TitleFull: Journal of Interactive Media in Education
              Type: main
ResultId 1