TPACK-based Professional Development for the AI Era: Fostering Pre-service Teachers' Acceptance of Generative AI in Mathematics Classrooms

TPACK-based Professional Development for the AI Era: Fostering Pre-service Teachers' Acceptance of Generative AI in Mathematics Classrooms | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article TPACK-based Professional Development for the AI Era: Fostering Pre-service Teachers' Acceptance of Generative AI in Mathematics Classrooms Shristi Shrestha, Jiyeong Yi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7622889/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract As Generative AI (GenAI) becomes more prevalent, the need to prepare pre-service teachers (PSTs) for its use is a critical challenge for mathematics teacher educators (MTEs). Yet, little is known about how to best foster PSTs’ adoption and critical use of GenAI in mathematics classrooms. This study addresses this gap by evaluating the impact of a 90-minute professional development workshop, grounded in the Technological Pedagogical Content Knowledge (TPACK) framework, on PSTs’ technology acceptance in mathematics education. A mixed-methods design was employed, using pre- and post-surveys based on an extended Unified Theory of Acceptance and Use of Technology (UTAUT) model for quantitative data and semi-structured interviews and workshop discussions for qualitative data. Quantitative analysis revealed statistically significant positive shifts in many aspects of technology acceptance, except for PSTs’ perceived risks of the technology. Qualitative analysis identified key facilitators to adoption, such as GenAI's utility for instructional efficiency, alongside significant barriers, including the lack of clear institutional guidance. The findings demonstrate that TPACK-based professional development opportunities can enhance PSTs’ responsible adoption of GenAI in mathematics education. This study provides actionable implications for MTEs on designing pedagogically grounded training that addresses GenAI's practical applications and ethical complexities in mathematics classrooms. Educational Psychology Artificial Intelligence and Machine Learning Generative AI Pre-service Teachers Mathematics Education Professional Development TPACK Technology Acceptance Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction The emergence of Generative Artificial Intelligence (GenAI) presents a significant pedagogical challenge for mathematics teacher educators (MTEs): how do we prepare pre-service teachers (PSTs) to critically and effectively use this emerging technology to support mathematical learning? The swift progress of GenAI is significantly transforming the educational landscape through the increasing integration of AI-powered resources into instructional practices. Research indicates that GenAI systems can help deliver differentiated instruction (Getenet, 2024 ) and promote student-centered learning (Song et al., 2025 ). Educators are adopting GenAI chatbots, like ChatGPT, alongside GenAI tools specialized for teaching, like Magic School AI, to streamline instructional responsibilities and provide tailored learning experiences for students (Baidoo-Anu & Ansah, 2023 ; Kasneci et al., 2023 ; Yilmaz & Yilmaz, 2023 ). In mathematics education, GenAI is recognized as a valuable resource for developing instructional content (Ellis & Slade, 2023 ), solving math problems (Guler et al., 2024 ), and providing personalized feedback (Li et al., 2023 ). PSTs have been found to be receptive to utilizing instructional materials created by GenAI, recognizing its benefits for planning lessons and designing assessments (Wang et al., 2024 ). However, these advantages do not come without concerns. PSTs have expressed apprehensions about GenAI’s reliability and its potential to undermine teacher autonomy (Yang & Appleget, 2024 ). These voiced reservations from PSTs, coupled with findings that many teacher educators lack the competencies to facilitate AI literacy for PSTs (Nyaaba & Zhai, 2024 ), highlight a significant gap in teacher preparation. Thus, it is imperative to develop structured training frameworks that equip PSTs with the knowledge and skills needed to thoughtfully and effectively incorporate GenAI tools in their teaching practices (Sánchez-Ruiz et al., 2023 ; Wang et al., 2024 ). Despite rising adoption of GenAI in mathematics classrooms, there is a significant research gap regarding the motivations that drive teachers to embrace these tools for teaching mathematics. Most studies investigating the acceptance of GenAI technology within educational settings have primarily focused on student usage patterns (Strzelecki, 2023 ; Faruk et al., 2023 ; Yilmaz et al., 2024 ). In contrast, there has been limited research exploring how pre-service and in-service teachers perceive the integration of GenAI into their pedagogical activities (Wang et al., 2024 ; Li, 2024 ). Given teachers' crucial role in determining technology adoption in classrooms, understanding the perspectives, concerns, and competencies of PSTs regarding GenAI is vital for developing effective training programs. Educators tend to find technology easier to use as they become more proficient (Mukuka & Alex, 2024 ). However, many teacher education programs lack hands-on opportunities for PSTs to develop and implement their GenAI knowledge in classroom settings (Macdowell et al., 2024 ). A lack of empirical research in the field mirrors this gap in teacher education programs. Although established technology acceptance frameworks like the Unified Theory of Acceptance and Use of Technology (UTAUT) model (Venkatesh et al., 2003 ) are frequently used to investigate technology adoption among educators (Gurer, 2021 ; Perienen, 2020 ), there is a dearth of research applying these models to understand how professional development (PD) shapes PSTs' adoption of GenAI, particularly within mathematics classrooms. Prior work establishes that PD rooted in pedagogical frameworks like the Technological Pedagogical Content Knowledge (TPACK) model (Mishra & Koehler, 2006 ) effectively enhances teacher confidence and willingness to adopt new technologies (Daher et al., 2021 ; Jaipal-Jamani & Figg, 2015 ). However, there is a scarcity of studies that systematically evaluate such learning programs for GenAI adoption. The UTAUT model serves as a robust foundation for studying technology adoption. However, it does not fully account for the unique challenges introduced by GenAI, namely user trust, perceived risk, and the pedagogical competencies and content knowledge required for effective classroom use. Therefore, this study employs an extended UTAUT framework that integrates the constructs of Trust, Perceived Risk, and TPACK. These additions are based on research identifying them as critical factors in the acceptance of AI-driven and educational technologies (Xiong et al., 2023 ; Yildiz & Arpaci, 2024 ). This enhanced model provides a more holistic lens for analyzing how PSTs perceive the adoption of GenAI tools and the underlying factors influencing their adoption decisions. This study addresses the gaps in the literature by designing and examining a TPACK-based PD workshop focused on building PSTs' technological and pedagogical knowledge to integrate GenAI tools into mathematics instruction. Using a mixed-methods approach, it quantitatively assesses the program's impact on PSTs' acceptance of GenAI in mathematics education and qualitatively explores the factors shaping their perceptions. Specifically, the research questions that guide this study are: How does a TPACK-based PD workshop impact PSTs' perceptions of GenAI’s adoption in mathematics classrooms? What factors influence PSTs’ willingness to adopt GenAI tools in mathematics education contexts? This study offers the research community an extended UTAUT framework, potentially providing a more nuanced lens for future GenAI adoption studies in educational contexts. For MTEs, the findings from the PD workshop may inform the design of pedagogically-grounded GenAI training programs. Ultimately, this work intends to inform strategies that enhance PSTs’ GenAI competencies and support them in navigating the pedagogical and ethical challenges of GenAI adoption in mathematics instruction. 2. Theoretical Perspectives The successful integration of technology in education begins with its acceptance by users (Appavoo, 2020 ). While several models exist to explain technology acceptance, this study builds upon the UTAUT model. UTAUT, introduced by Venkatesh et al. ( 2003 ), is a comprehensive model that explains a user's intention to adopt a technology. The model posits that four core constructs directly influence user acceptance: Performance Expectancy (PE): The degree of belief that using the technology will enhance job performance. Effort Expectancy (EE): The perceived ease of using the technology. Social Influence (SI): The social pressure an individual feels to use the technology. Facilitating Conditions (FC): The perceived availability of organizational and technical resources to support the technology's use. In the UTAUT model, the first three constructs—PE, EE, and SI—are direct determinants of Behavioral Intention (BI), which is a user's intention to use the technology. This intention, along with FC, then predicts the actual Use Behavior (UB), i.e., the genuine adoption of the technology. The model also includes moderating variables such as age, gender, experience, and voluntariness of use that influence relationships between the constructs, as demonstrated in Fig. 1 . UTAUT has been validated in various education research studies (Aytekin et al., 2022 ; Yee et al., 2021) and has been widely used to assess the acceptance of AI technologies like prediction and recommendation systems (Menon & Shilpa, 2023 ). However, the unique aspects of GenAI for educators require an extension of the UTAUT model. GenAI introduces significant risks, including data bias, misinformation, and privacy violations (Weidinger et al., 2023 ), which undermine user trust (Menon & Shilpa, 2023 ). Trust is crucial for enhancing user acceptance, while perceived risks can hinder it (Ajenaghughrure et al., 2021 ; Choung et al., 2022 ). Thus, their inclusion in UTAUT is critical for understanding GenAI adoption, an approach adopted and validated by Xiong and colleagues ( 2023 ). Similarly, pedagogical competence is a critical determinant of technology adoption for educators. TPACK, introduced by Mishra and Koehler ( 2006 ), illustrates the vital interplay among technology, pedagogy, and content knowledge for effective technology integration in education. Research (Kiyici & Övez, 2021 ; Mayer & Girwidz, 2019 ) has consistently established the significant influence of TPACK on PSTs’ and mathematics teachers’ technology acceptance and their perceptions of a tool’s usefulness and ease of use. Additionally, Wang et al. ( 2024 ) found that PSTs’ self-perceived TPACK enhanced their acceptance of GenAI by improving their perceptions of its performance and effort expectancy. Therefore, this study employs an extended UTAUT framework that integrates these three key constructs—Trust (TR), Perceived Risk (PR), and TPACK—to create a more nuanced understanding of GenAI acceptance among PSTs, as shown in Fig. 2 . This extension deliberately omits the original model's moderating factors to maintain a sharper focus on the central constructs most relevant to the research questions. 3. Literature Review This section synthesizes literature to establish the foundation for this research. It first explores the affordances and challenges of integrating GenAI in mathematics education, then examines key factors influencing technology acceptance in educational contexts, and concludes by reviewing the role of PD in enhancing technology adoption among PSTs. 3.1 GenAI in Mathematics Education GenAI’s integration into mathematics education offers significant potential but presents considerable challenges. GenAI supports educators in generating course materials and lesson plans (Bazelais et al., 2024 ; Ellis & Slade, 2023 ). For students, it enhances learning by assisting in solving problems, explaining concepts, and providing personalized feedback and hints (Guler et al., 2024 ; Dasari et al., 2024 ; Li et al., 2023 ). Song et al. ( 2025 ) found that GenAI-powered agents can foster student-centered learning by improving engagement and conceptual understanding. Gattupalli et al. ( 2023 ) found that PSTs sometimes prefer GenAI’s step-by-step guidance over human-crafted hints in online math learning. Furthermore, GenAI can be used to differentiate instruction (Getenet, 2024 ), create engaging tasks (Sánchez-Ruiz et al., 2023 ), and complement traditional tools like Computer Algebra Systems (CAS) (Matzakos et al., 2023 ), with research highlighting its promise in teaching specific topics like geometry (Wardat et al., 2023 ) and its positive reception among students (Sánchez-Ruiz et al., 2023 ). However, these advantages are contested by significant ethical and pedagogical concerns that warrant careful consideration. Key challenges include risks to academic integrity (Rahman & Watanobe, 2023 ), the potential spread of misinformation due to hallucinations (Athaluri et al., 2023 ), the reinforcement of biases (Gross, 2023 ), and the likelihood of exacerbating the digital divide (Chan & Hu, 2023 ). In the context of mathematics education, particular limitations involve questionable computational accuracy when compared to specialized CAS tools (Matzakos et al., 2023 ), an overreliance on AI by students for answers, and pedagogical gaps such as the lack of adaptability to diverse learning styles (Sánchez-Ruiz et al., 2023 ). These issues indicate that GenAI should complement, rather than replace, traditional teaching methods (Dasari et al., 2024 ; Ellis & Slade, 2023 ). Thus, effective integration of GenAI in mathematics classrooms hinges on targeted training for teachers to navigate these limitations and leverage GenAI’s full potential. 3.2 Technology Acceptance in Educational Contexts Technology acceptance explains how and why individuals embrace new technologies. The acceptance of technologies in education is a widely researched area, frequently analyzed through theoretical frameworks like the UTAUT model (Venkatesh et al., 2003 ) and its variants. This section presents literature to help understand the factors influencing the acceptance of GenAI and other technologies in general and math-specific educational settings. GenAI acceptance research in higher education has shown that several key factors drive students' intention to use these tools. Across multiple studies, performance expectancy—the perceived usefulness of technology—consistently emerges as a primary determinant of adoption (Bazelais et al., 2024 ; Faruk et al., 2023 ; Strzelecki, 2023 ; Wang et al., 2024 ). Beyond utility, student acceptance is also influenced by habit (Strzelecki, 2023 ), hedonic motivation, i.e., enjoyment (Strzelecki, 2023 ), and contextual factors like perceived humanness and novelty value (Faruk et al., 2023 ). Similarly, Strzelecki ( 2023 ) identified facilitating conditions (e.g., technical support and resources) as a critical driver. From students’ psychological standpoint, personality traits such as openness can positively affect usage, while neuroticism may act as a barrier (Faruk et al., 2023 ). Researchers have also developed standardized instruments like the Generative Artificial Intelligence Acceptance Scale (GAIAS) (Yilmaz et al., 2024 ) to better measure these constructs for educational contexts, validated with a sample of 627 university students. While student adoption has been widely studied, research on educators remains less common. A notable study by Wang et al. ( 2024 ) examined PSTs’ intention to use GenAI, finding that performance expectancy and social influence were the strongest predictors. Their extended UTAUT model also incorporated technology self-efficacy and TPACK, finding that higher levels in these areas increased confidence and reduced GenAI anxiety. Interestingly, some studies have found constructs like effort expectancy, i.e., the perceived ease of use, and facilitating conditions, to have no significant impact on behavioral intention to use GenAI in learning and teaching (Bazelais et al., 2024 ; Wang et al., 2024 ). This suggests that for some users, the perceived benefits of GenAI may outweigh the challenges of use or the need for support. For mathematics teachers and PSTs, perceived usefulness and performance expectancy consistently predict their intention to use technology (Al-zboon et al., 2021 ; Gurer, 2021 ; Philemon, 2022; Wong, 2015 ). Effort expectancy was also found to be an important factor influencing adoption (Gurer, 2021 ; Perinen, 2020; Philemon, 2022). Additionally, facilitating conditions and social influence are often cited as powerful influences on technology adoption (Wong, 2015 ; Gurer, 2021 ; Al-zboon et al., 2021 ). A study by Perienen ( 2020 ) noted that despite recognizing technology's pedagogical value for mathematics instruction, teachers' actual integration was minimal, with perceived ease of use and facilitating conditions being the primary factors impacting use. This highlights a persistent need for PD focused on pedagogical integration of technology in classrooms. Furthermore, factors such as teacher attitudes, pedagogical beliefs, and technology self-efficacy are crucial in predicting technology acceptance (Wong, 2015 ; Gurer & Akkaya, 2022 ; Li, 2024 ), as is the alignment of the technology with instructional goals, i.e., task-technology fit (Philemon et al., 2022 ). As GenAI tools enter mathematics education, early research shows positive perceptions among US college students (Li et al., 2024 ), but teacher adoption remains complex. A recent study on Chinese primary mathematics teachers by Li ( 2024 ) found that positive teacher attitudes significantly impacted GenAI adoption. While the literature provides a solid foundation for understanding technology acceptance in educational contexts, a significant gap persists at the intersection of GenAI technology, teachers, and mathematics education. Research on GenAI has predominantly centered on students, while studies in mathematics education have only begun exploring GenAI. Consequently, the factors influencing the acceptance and use of GenAI tools specifically among pre-service mathematics teachers remain largely unexplored. 3.3 Enhancing Technology Acceptance of PSTs with PD The literature consistently emphasizes the crucial significance of training programs and PD in facilitating technology integration into instructional practices (Mistretta, 2005 ; Ndlovu et al., 2020 ). While a plethora of studies examine the factors influencing technology acceptance among educators, there remains a noticeable gap in research focused on interventions designed to improve this acceptance (Kale, 2018 ). In this vein, Daher and colleagues ( 2021 ) conducted a noteworthy investigation into the effectiveness of targeted PD in enhancing technology acceptance. Their study evaluated how the preparation of PSTs for utilizing digital tools in mathematics and science instruction influenced their acceptance of these technologies. The preparation model centered on developing TPACK, emphasizing technical proficiency and pedagogical understanding of digital tools. The intervention program incorporated collaborative discussions, the creation of digital content, and reflective practices regarding implementing these tools in authentic classroom environments. Analysis using paired sample t-tests revealed significant increases in pre- and post-intervention scores across several dimensions, including perceived ease of use, perceived usefulness, attitudes, intent to use, actual usage, and self-efficacy. Notably, anxiety levels did not exhibit a significant change. In summary, the teacher preparation program effectively transformed PSTs' attitudes and perceptions, enhancing several aspects of technology acceptance regarding digital tools. Another study by Özbek and colleagues ( 2023 ) also demonstrates that educational interventions focused on new technology positively impact their acceptance among PSTs. The researchers indicated that various forms of engagement with tool-specific information are crucial in promoting this acceptance. Specifically, their investigation revealed that both PST groups, one that engaged in hands-on learning tasks with a new digital tool and the other that read a blog post about it, experienced an increase in their intention to incorporate the technology into their lesson plans. Furthermore, Yang and Appleget ( 2024 ) underscored the importance of structured integration of GenAI within teacher education programs. In their study, PSTs utilized GenAI tools to generate and assess read-aloud questions during a literacy methods course, applying their pedagogical and content knowledge in a practical learning setting. The findings indicated that 91% of PSTs recognized GenAI as a valuable instructional tool, appreciating its efficiency. However, the PST also expressed concerns regarding the possible adverse effects of the technology on teacher agency and creativity, as well as the questionable reliability of the technology. A positive correlation was identified between PSTs' engagement with GenAI and their intentions to incorporate it in future teaching, highlighting the critical need for AI literacy and pedagogical adaptation within teacher preparation curricula. Overall, the relevant literature suggests that educational interventions are vital in enhancing PSTs' acceptance of technology. By integrating insights from the studies mentioned, we posit that professional learning opportunities that offer exposure to the application of GenAI in mathematics education, combined with interactive activities, could significantly augment PSTs' acceptance of this technology. 4. Methods 4.1 Research Design This study employed a sequential mixed-methods design (Creswell, 2021 ) to comprehensively understand PSTs’ attitudes toward using GenAI in mathematics education. This approach has been known to be particularly effective in technology acceptance research in educational settings (Creswell & Clark, 2018; Venkatesh et al., 2013 ), as it integrates the strengths of quantitative data to measure outcomes with qualitative data to explore the underlying processes driving them. As illustrated in Fig. 3 , the initial quantitative phase addressed Research Question 1 (RQ1) by assessing the effects of an intervention workshop on PSTs' perceptions. The subsequent qualitative phase addressed Research Question 2 (RQ2), exploring the underlying factors that shaped these perceptions. This sequential structure ensured that the quantitative data informed the qualitative data collection, particularly for designing the semi-structured interview guide. 4.2 Context and Participants This study was conducted in two teacher education courses at a large public university in the Midwest United States. One course focused on mathematics methods, consisting of 12 students, and another was an educational technologies course with 15 students in one section and 14 students in another. The first author facilitated three 90-minute workshops, one in the mathematics methods course and two in each section of the learning technologies course. All 41 PSTs enrolled in these courses participated in the workshops as part of their regular coursework and were invited to participate in the research. A total of 28 PSTs (8 male, 20 female) consented to participate, forming a convenience sample. This sample included various academic levels: six sophomores, 12 juniors, eight seniors, and two master’s students. No significant demographic differences were observed between the 28 participants and the 13 non-participants, suggesting the study sample reasonably represented the students enrolled in the courses. Following the workshop, all 28 participants were invited via email to volunteer for a follow-up interview to discuss their experiences in more depth. The first five PSTs who responded and aspired to serve as mathematics educators were selected for the interviews. This purposive sub-sample included three females and two males, representing a range of academic levels: one junior, three seniors, and one master's student. The group comprised four secondary mathematics PSTs and one elementary education PST pursuing a secondary mathematics endorsement. This selection provided a focused perspective from participants whose career paths aligned directly with the study's focus on mathematics education. 4.3 TPACK-based PD Workshop The intervention consisted of a 90-minute workshop to enhance PSTs’ competence and willingness to integrate GenAI in mathematics instruction. The workshop's design was grounded in the TPACK-based Professional Learning Design Model (TPLDM), a framework developed by Figg and Jaipal-Jamani ( 2012 ; 2013 ; 2015 ). The TPLDM was selected because it moves beyond traditional, technocentric PD by providing a structured approach that systematically integrates technological, pedagogical, and content knowledge. Research has demonstrated the model's effectiveness in building TPACK competencies among diverse educators, from university faculty to K-12 teachers (Jaipal-Jamani et al., 2018 ; Tai, 2015 ). The 90-minute workshop guided the PSTs through a carefully structured sequence of interactive activities organized around the four key components of the TPLDM framework, described below. 4.3.1 Modeling a Technology-Enhanced Activity Type The workshop started by placing the PSTs in a scenario-based simulation where they assumed the role of a sixth-grade student tackling ratio practice problems on Khan Academy, an online educational platform. During this modeled activity, the PSTs worked through several ratio problems both with and without the support of Khanmigo, a GenAI chatbot integrated into Khan Academy. Khanmigo was chosen for this demonstration because of the widespread use and recognition of Khan Academy among school districts in the U.S., making it a relevant and relatable example for PSTs. During the 2023–2024 school year, Khan Academy had nearly 975,800 licensed users across 577 U.S. school communities (Khan Academy, 2024 ). The simulation employed a sequence of images that followed a narrative, allowing PSTs to observe and critically assess the tool’s ability to provide instant feedback, scaffold problem-solving strategies, and tailor instruction to meet student needs. For instance, during the activity demonstrated in Fig. 4 , some PSTs raised concerns that while Khanmigo effectively prompted the student to identify the correct numbers needed for a ratio problem (five turtles and eight total animals), it immediately provided the final answer. Instead of guiding the student to form the ratio themselves, the chatbot stated, "So, the ratio of turtles to total animals is 5 for every 8." After the simulation, a volunteer PST participated in a live, hands-on session with Khanmigo, mimicking a student's real-time navigation of the tool. The PSTs engaged with Khanmigo in various manners—they used the chatbot's suggested prompts, communicated with it using language typical of a sixth grader (including slang and spelling mistakes), and exhibited behaviors resistant to learning. This interactive demonstration enabled PSTs to closely examine how real-time interactions with a GenAI tutor resemble the experiences of their perceived impression of a sixth-grade learner. Thus, this activity, which lasted approximately 25 minutes, contributed to PSTs' understanding of the complex relationship between GenAI tools (Khanmigo), pedagogical strategies (scaffolding and differentiation), and math content (ratios). 4.3.2 Integrating Pedagogical Dialogue After the modeling exercise, PSTs took part in timed discussions on instructional implications of GenAI, particularly examining the relationship among pedagogy, mathematical content, and GenAI technology. The discussions centered on two key questions: (1) How can Khanmigo facilitate differentiation for mixed-ability classrooms? (2) What could be some drawbacks to students engaging with GenAI bots like Khanmigo? As illustrated in Fig. 5 , the PSTs initially answered these prompts on their own for approximately one minute using an interactive educational tool, Wooclap. This was followed by two-minute small group discussions, and then a whole-class conversation that took about two minutes to integrate various perspectives. This stage specifically focused on the pedagogical aspect of the TPACK framework, encouraging PSTs to reflect on how technology and teaching methods intersect and affect student learning in mathematics education. 4.3.3 Developing Technical Skills through Short Tool Demonstrations In this part, PSTs were asked to individually respond, via Wooclap, to the prompt, “How are GenAI tools being used by teachers for mathematics instruction?” This reflective exercise aimed to elicit their perceptions and knowledge about using GenAI tools in mathematics teaching. Then, the first author showed live demonstrations for about 40 minutes to showcase the instructional uses of various GenAI tools: ChatGPT, Microsoft Copilot, Google Gemini, and Anthropic Claude Sonnet. The workshop illustrated specific instructional applications pertinent to mathematics education, which included: Ideating interactive story-based problems for specific learning standards and goals Adjusting a middle school-level percentage problem about discounts to an elementary-level addition/subtraction problem Simplifying a mathematical word problem to improve the question’s readability Generating hints to provide students with scaffolding to solve math problems Generating a graph that depicts the story of the race between the tortoise and the hare via piecewise linear functions Each instructional application was showcased using at least two distinct GenAI tools (e.g., ChatGPT and Google Gemini), prompting comparisons of their outputs based on the same prompt. Through these demonstrations, the PSTs witnessed how GenAI tools could support lesson planning, differentiation, and instructional scaffolding, providing them with essential technical skills to implement these technologies in their future classrooms. Consequently, PSTs were provided fundamental technical skills, effectively merging technological and content knowledge crucial for real-world classroom applications. 4.3.4 Applying TPACK to Design Their Own Task The workshop concluded with PSTs applying their integrated technological, pedagogical, and content knowledge by independently creating instructional materials using GenAI. The PSTs selected one or more GenAI tools that had been previously showcased to develop materials that were aligned with specific mathematics learning objectives. The tasks drew on earlier demonstrated use cases, which allowed the PSTs to transform the knowledge acquired during the workshop into tangible teaching resources. Figure 6 illustrates the submitted work of a participating PST during this exercise, where the individual employed ChatGPT to generate hints for a mathematical word problem. This design activity enabled the PSTs to synthesize their learning by applying GenAI to enhance pedagogical strategies within the context of mathematics, thereby enabling an advancement in their TPACK competencies. 4.4 Data Collection and Analysis Pre- and post-surveys were administered via Qualtrics immediately before and after the workshop. The survey instrument included 29 items (see Appendix) measuring the nine constructs of the extended UTAUT model (PE, EE, SI, FC, TR, PR, TPACK, BI, and UB). The instrument was created by adapting items from previously validated instruments in recent, relevant literature (Ning et al., 2024 ; Xiong et al., 2023 ; Yildiz & Arpaci, 2024 ; Yilmaz et al., 2023) to ensure validity of the measured constructs. The instrument used a 7-point Likert scale, instead of five, to improve the scale’s sensitivity (Preston & Coleman, 2000) and reduce bias (Garland, 1991 ). Initial descriptive statistics were performed to review pertinent variables. Specifically, mean, median, and standard deviation (SD) were calculated to assess central tendency and variability. Among the 28 participating PSTs, data from three were excluded due to incomplete survey responses, resulting in a final sample of 25 for inferential analysis. Composite scores for each model construct were computed by averaging their respective item scores, thus providing an overall assessment of each construct both pre- and post-workshop. The internal consistency and reliability of the adapted instrument were confirmed with the current study's data, with Cronbach’s alpha coefficients (α) for each construct, with a reliability threshold set at 0.7. Before conducting inferential analyses, the Shapiro-Wilk test assessed the normality of each composite score distribution. A paired-samples t-test (for normally distributed data) or a Wilcoxon signed-rank test (for non-normally distributed data) was used to compare pre- and post-survey scores to determine the intervention workshop's impact, with a significance threshold set at p < 0.05. Qualitative data were gathered from two primary sources, i.e., semi-structured interviews and workshop responses, to provide a deep and contextualized understanding of PSTs’ perceptions. Following the workshop, in-depth semi-structured interviews, lasting 40 minutes to an hour, were conducted with five PSTs specializing in mathematics education. The interviews were audio-recorded and transcribed verbatim. The interview protocol was aligned with the study’s theoretical framework to explore participants’ attitudes towards adopting GenAI in educational contexts. Participants digitally submitted individual written responses to discussion prompts and images of hands-on tasks during the workshop via the Wooclap platform. For the three workshops’ discussion prompts, 24 participating PSTs submitted their thoughts, collecting a total of 88 responses. The number of individual responses for each question was as follows: The first question, "How can Khanmigo facilitate differentiation for mixed-ability classrooms?" received 24 responses. The second question, "What could be some drawbacks to students engaging with GenAI bots like Khanmigo?" received 29 responses. Finally, the third question, "How are GenAI tools being used by teachers for mathematics instruction?" received 35 responses. An integrated thematic analysis (Braun & Clarke, 2006 ) was conducted to synthesize findings from all qualitative data sources. The analysis, managed in the Dedoose software, followed a multi-phase coding process to ensure a robust and comprehensive interpretation. An initial coding phase was conducted using a deductive approach, where the constructs of the extended UTAUT model served as a priori parent codes. This step ensured that the analysis aligned with the study’s theoretical framework. Then, a second inductive coding phase was conducted within and across the UTAUT constructs to identify emergent themes and patterns. This allowed for capturing nuanced perspectives not fully encompassed by the initial framework. The analysis was primarily conducted by the first author, so many strategies were employed to ensure the trustworthiness and rigor of the findings. The qualitative findings from the interviews and workshop responses were systematically compared with each other and against the quantitative survey data. This methodological triangulation was used to identify areas of convergence, where both data types told the same story, and expansion, where the qualitative data provided a deeper explanation for the quantitative results. This process created a more valid and holistic interpretation of the PSTs' perceptions. Peer debriefing served as the primary strategy for ensuring analytical rigor. The first author regularly met with the second author to review the codebook, discuss analytical decisions, and challenge emerging interpretations. This critical dialogue helped validate the thematic structure and minimize potential researcher bias. A detailed audit trail was maintained, documenting every step of the analytical process, from creating and revising the codebook to developing the final themes. 5. Results 5.1 Shifts in PSTs’ Technology Acceptance of GenAI in Mathematics Classrooms To evaluate the impact of the PD workshop, shifts across eight constructs of the extended UTAUT model, except UB, were measured. The pre-workshop UB scores (M = 5.92, SD = .81) captured a baseline of the PSTs’ pre-existing use of GenAI tools. Since the pre- and post-surveys were administered immediately before and after the workshop, no change was anticipated in the use behavior of PSTs. So, inferential analysis was not performed on the UB scores. Internal consistency and reliability of the adapted instrument were confirmed with the study's data from pre- and post-workshop surveys, with Cronbach’s alpha coefficients for constructs showing acceptable (.70 ≤ α < .80), good (.80 ≤ α < .90), or excellent (α ≥ .90) reliability, as shown in Table 1 . Table 1 Shifts in Pre- and Post-workshop Survey Scores for Extended UTAUT Constructs Paired Samples t-test Construct Pre-workshop Post-workshop Mean Score Difference (M) Test statistic (t) p-value (p) Cronbach’s Alpha (α) Mean (SD) Cronbach’s Alpha (α) Mean (SD) EE 0.89 5.33 (1.06) 0.86 5.70 (0.87) 0.37 -2.47 0.0212 SI 0.80 4.57 (1.00) 0.75 5.05 (0.87) 0.48 -2.41 0.0239 TR 0.79 5.09 (1.13) 0.95 5.40 (1.09) 0.31 -2.35 0.0272 PR 0.79 3.96 (1.37) 0.83 3.73 (1.40) -0.23 1.80 0.0842 TPACK 0.86 5.29 (0.99) 0.79 5.85 (0.79) 0.56 -2.99 0.0052 BI 0.78 5.13 (0.90) 0.79 5.67 (0.80) 0.55 -4.67 0.0001 Wilcoxon Signed-rank Test Construct Pre-workshop Post-workshop Median Score Difference ( M ) Test Statistic (W) p-value (p) Cronbach’s Alpha (α) Median (SD) Cronbach’s Alpha (α) Median (SD) PE 0.85 5.33 (0.93) 0.90 6.00 (0.73) 0.67 9.50 0.0003 FC 0.82 4.67 (1.21) 0.83 5.67 (1.05) 1.00 27.50 0.0199 The descriptive and inferential statistics, summarized in Table 1 , reveal a significant positive shift in PSTs’ perceptions of GenAI following the workshop. Participants reported statistically significant improvements in all constructs measured, except PR (p > .05). The most substantial changes were observed in constructs directly related to the PSTs' intention to adopt GenAI (BI; M = .55, t = -4.67, p = .0001) and perception of available resources and support to use GenAI in educational contexts (FC; M = 1.00, W = 27.5, p = .0199). This was complemented by highly significant gains in PE, indicating that after the workshop, PSTs more strongly believed that GenAI would help achieve their academic and instructional goals ( M = .67, W = 9.5, p = .0003). Furthermore, participants’ confidence in integrating GenAI, pedagogy, and math knowledge (TPACK) grew significantly (M = .56, t = -2.99, p = .0052). Significant improvements were also found in PSTs’ perceived ease of use of GenAI (EE; M = 0.37, t = -2.47, p = 0.0212), perceived influence from important others to adopt GenAI in educational contexts (SI; M = .48, t = -2.41, p = .0239), and trust on the technology (TR; M = .31, t = -2.35, p = .0272). In contrast, the decrease in PR was not statistically significant (M = -0.23, t = 1.80, p = .0842), suggesting that while other perceptions became more positive, PSTs’ sense of risk associated with using GenAI did not decrease significantly. 5.2 Factors Influencing PSTs’ Willingness to Adopt GenAI The integrated thematic analysis of interview and workshop data revealed five overarching themes. These themes synthesize the constructs of the extended UTAUT model into a holistic narrative of the facilitators and barriers shaping the PSTs’ perspectives. The themes are supported by excerpts from the interview and workshop, with participants referred to by pseudonyms. 5.2.1 Theme 1: GenAI as a Collaborative Partner for Efficiency and Creativity This was the most dominant theme, identified by all five interviewees and heavily supported by the workshop data, where 57 out of 88 submitted responses were related to GenAI's utility. This theme encapsulates PSTs' views on GenAI’s perceived usefulness (PE) and their current use behavior (UB), framing GenAI not just as a tool, but as a partner that enhances their professional practice. PSTs consistently described using GenAI to offload the repetitive, time-consuming aspects of teaching. This focus on efficiency was detailed in the workshop. In response to a question about how they think math teachers are using GenAI, participants identified the rapid generation of instructional materials as the key benefit, specifically highlighting its ability to create problems (11 responses), lesson plans (7 responses), and differentiated materials (7 responses). This sentiment was perfectly captured during an interview with Liberty, a junior in elementary education. She explained how she planned to use these tools to maintain a healthy work-life balance in her future career. She saw GenAI as a practical assistant that would "help me stay at school within my contracted hours." Beyond simple efficiency, PSTs valued GenAI as a creative collaborator that could enhance the quality of their instruction. Caleb, a master's student, provided a vivid description of this dynamic process during his interview, sharing that he does not ask GenAI for a finished product but instead engages in a brainstorming session. He explained, "You can use it almost like a whiteboard; you just throw an idea to it, and it will throw an idea back, and you just keep bouncing back and forth until you get what you want." This collaborative potential was also seen in its application for students, a point that emerged from the workshop discussion on differentiation. Participants envisioned tools like Khanmigo acting as an intelligent, personalized tutor capable of "creating different tiers of problems for different tiers of learners" and providing instant scaffolding that "meets [students] where they are at with their knowledge." By offering "guided help" for struggling learners and novel challenges for advanced ones, PSTs saw GenAI as a direct partner in fostering a more creative and differentiated learning environment. 5.2.2 Theme 2: Navigating a Learning Curve with Cautious Engagement This theme, central to the experience of all five interviewees, combines PSTs’ perceived ease of use (EE) and trust in the GenAI technology (TR) and their developing TPACK. It captures the understanding that while GenAI is accessible, leveraging it effectively is a learned skill that requires critical oversight. Caleb shared that the quality of the output is the user's responsibility, not the tool's. He explained, "GenAI is very dumb. It's as smart as you are. If you give it a well-detailed design prompt, it will give you a well-detailed answer. But if you give it a generic prompt, it'll give you a generic response." This sentiment was echoed by Amanda, a senior, who noted that getting a useful output "takes some practice" and that "you have to learn how to feed it the information so that you get what you want." This reframes ease of use not as a static property of the tool, but as a skill the user develops over time. This cautious engagement is rooted in a situational trust in the technology. Liberty described a practical workflow that perfectly illustrates this balance. While she doesn’t "trust it 100%," she confidently uses it for creative tasks like generating story problems. She then applies her own expertise, explaining that "if it doesn't give me exactly what I'm looking for, I can use what it gave me. I switch it up and do what I'm looking for." This sentiment was shared by Jade, another senior, who trusts GenAI to explain a concept but has "trust issues with it" when asking it to solve a specific math problem. These examples show PSTs are already developing a nuanced TPACK, learning to discern which tasks are appropriate for AI and which require their own critical evaluation and expert verification. 5.2.3 Theme 3: Managing Risks of Inaccuracy and Dependency A significant concern, raised by all five interviewees and dominating the workshop discussion on drawbacks of Khanmigo (29 responses), was the inherent risk associated with GenAI (PR). This theme combines perceived risk of GenAI technology with skepticism around its usefulness, highlighting the ethical and pedagogical dilemmas PSTs foresee. The most frequently cited risk, mentioned in 22 of the 29 workshop responses and by all five interviewees, was the potential for student over-reliance leading to reduced cognitive engagement. PSTs expressed concerns that when students repeatedly turn to AI for instant answers instead of struggling with problems independently, they fail to engage in the deeper cognitive processing necessary for true understanding. Workshop participants voiced a fear that students would learn to manipulate the AI (11 responses), noting they might "coax out an answer [from AI] without them trying the problem" or use it for "just giving the answer instead of giving an explanation as to how they can solve a specific problem." This superficial approach, as Serenity, a workshop participant, stated, leads to students “getting the answer without showing interest in learning the process." Compounding this issue was the risk of students learning from inaccurate AI-generated responses. Caleb, during his interview, elaborated on this with a powerful cognitive example, warning that if a student learns a concept incorrectly from an AI that confidently gives a wrong answer, "it's going to be very hard to get rid of it," underscoring the high stakes of misinformation in foundational learning. Beyond these pedagogical concerns, all interviewed PSTs identified concerns around data protection and student privacy. Amanda articulated this particularly well, worrying about the unseen data implications of using these tools with students. She explained, "You have to be careful what you ask them to do on their school computers. You don't know what pulls information. With FERPA and the fact that they're minors, that's always like a concern because, me as an adult, it's that fine line where I still need to protect their data." This demonstrates a sophisticated awareness of professional responsibilities and the potential ramifications of uncritical GenAI adoption. 5.2.4 Theme 4: Navigating an Ambiguous Professional Landscape This theme, a major point of frustration for all five interviewees, encompasses the elements of influence PSTs receive from important others (SI) and their perception of available resources and support (FC). It highlights how PSTs are navigating a simultaneously encouraging and unsupportive professional landscape, receiving conflicting advice and lacking clear institutional guidance on how to proceed with GenAI. PSTs are receiving explicit advocacy from within their teacher education programs. They acknowledged that some courses have made efforts to introduce GenAI through class activities like a "Stump the AI" exercise, where AI chatbots are challenged with questions they are likely to get wrong. Two PSTs shared experiences of assignments that required them to use GenAI to create lesson plans and then critique the AI-generated outputs. However, four out of the five interviewees described this exposure as superficial. Logan, a senior, expressed a desire for more structured and practical guidance: "They told us that it exists, but they should have been teaching us how to use it to create more innovative lesson plans...so that it's not something that we have to figure out by ourselves later down the road when we're swamped with lesson planning." This gap in practical training is compounded by a confusing lack of consensus from the wider academic and professional communities. Amanda, a PST co-majoring in Psychology, vividly described this institutional inconsistency, noting how her professors' views were starkly divided by department. She explained that the Psychology department wants to "pretend like it is not even a thing," while the Education department is "trying to figure out how to use it." This creates a confusing environment where PSTs are simultaneously encouraged and discouraged from using the same technology. This ambiguity extends to the K-12 environment they are preparing to enter. Jade stated, "I don't wanna be the only one who's using it [GenAI] or not using it and then finding out everyone else uses it or doesn't use it. Because it's something that a lot of people are very strong for, and then some don't want you using it at all. And just making sure I find that common level of what is expected of me from higher up people and districts." This highlights that a major barrier to adoption isn't just personal skill or belief, but the absence of a consistent professional framework for AI integration. 5.2.5 Theme 5: A Commitment to Proactive and Pragmatic Integration The final theme reflects the PSTs' ultimate stance, combining their pre-existing GenAI usage (UB), developing TPACK, and strong intentions to use the technology (BI). Voiced by all five interviewees, this theme shows how PSTs are not waiting passively for guidance but are instead proactively cultivating their own professional approach to using GenAI. PSTs are already integrating GenAI into their workflows as a strategic and pragmatic tool. Liberty provided an excellent example, explaining that she uses Magic School AI to "quickly create presentations, lesson plans, and write professional emails." This proactive use is coupled with a self-aware development of their TPACK. PSTs were consciously building the skills to use these tools effectively while recognizing their limitations. Jade expressed this evolving confidence: "Not fully confident right now, especially because I don't know of all the AI websites that are out there...I may have to just redo a whole bunch of things, such as learning what is fully reliable." During the interview, Logan recounted his experience using GenAI the previous summer, when he co-taught a summer school program at an urban middle school as part of his university's outreach initiative. He stated, "One of the lessons I taught this summer was a statistics unit, and I used ChatGPT and Copilot to design simple experiments for my students. I can see myself using these tools more for similar tasks in the future." This illustrates that PSTs are committed to developing their strategies for integrating GenAI. 6. Discussion This study investigated the impact of a TPACK-based PD workshop on PSTs’ acceptance of GenAI in mathematics classrooms. The quantitative findings demonstrate that the intervention led to statistically significant improvements across seven extended UTAUT constructs, except PR. This aligns with prior research indicating that TPACK-based PD effectively enhances technology acceptance (Daher et al., 2021 ). A primary outcome of the workshop was a significant boost in PSTs’ perceived utility and capability to use GenAI. The highly significant increase in PE suggests the workshop illustrated GenAI’s practical benefits to the PSTs, a finding consistent with other intervention studies (Daher et al., 2021 ; Özbek et al., 2023 ). This was reflected in the qualitative data, where PSTs framed GenAI as a "collaborative partner" for enhancing efficiency and creativity. These perceptions align with findings from Yang and Appleget ( 2024 ), where PSTs recognized GenAI as a useful teaching tool. However, this optimism did not come without criticism. PSTs remained skeptical about GenAI’s utility for complex mathematical reasoning and problem solving, corroborating research highlighting its computational limitations (Matzakos et al., 2023 ). This enhanced sense of utility and a significant increase in self-reported TPACK created a strong foundation for adoption. The pre-workshop survey yielded a mean TPACK score of 5.29, indicating that many participants already felt confident in integrating GenAI with pedagogy and content before the intervention. The statistically significant increase in post-workshop scores suggests that the intervention expanded PSTs’ knowledge, boosting their confidence. Direct engagement with GenAI tools and structured pedagogical discussions likely facilitated this growth, allowing PSTs to explore AI's potential in their teaching contexts. These results align with prior research indicating that TPACK-based PD enhances teachers’ perceived ability to integrate technology into instruction (Dalal et al., 2017 ; Meletiou-Mavrotheris & Paparistodemou, 2024 ). However, qualitative insights also revealed that some participants felt uncertain about implementing GenAI tools in real classrooms and expressed a need for more practical experience. This highlights the importance of continued exposure and structured practice to solidify PSTs’ skills and confidence in leveraging GenAI, as Bae et al. ( 2024 ) suggested. The significant increase in EE suggests that the workshop lowered barriers to entry for using the technology. The live demonstrations and hands-on tasks could have improved the PSTs’ perceived ease of use. During the interviews, PSTs shared that moving beyond basic use of GenAI requires skills and practice to leverage the technology’s full potential. One of the skills noted was prompt engineering, validating calls to formally teach prompting strategies in higher education (Lee & Palmer, 2025 ). One of the most insightful findings of this study is the dynamic between trust and risk. While PSTs’ trust in GenAI (TR) for educational tasks significantly increased, their perception of its risks (PR) did not significantly decrease. This outcome can likely be attributed to the workshop's design, which facilitated a discussion of GenAI's risks without providing explicit strategies for their mitigation. By making potential threats more salient and concrete, the workshop did not alleviate fear but rather transformed it into a more informed, critical awareness. This is reflected in the modest post-workshop TR score (M = 5.40) and the qualitative theme of "cautious engagement," which indicates that PSTs trust GenAI for specific tasks (e.g., creative brainstorming) while remaining wary of its limitations (e.g., for high-stakes mathematical problem-solving). The interviewed PSTs emphasized the need for human oversight, a concept extensively advocated in the literature (Deng & Joshi, 2024 ; Razmerita, 2024 ). Ultimately, the PSTs’ persistent risk perception reflects a sophisticated professional understanding of GenAI’s potential for misinformation and student dependency, a well-documented concern (Hsu et al., 2024 ; Rahman & Watanobe, 2023 ). This understanding is a more desirable outcome for teacher education than uncritical acceptance. Additionally, the findings highlight the critical role of the institutional context. The significant increases in SI and FC indicate that the workshop, which was delivered as a part of the PSTs’ teacher education courses, was perceived as a form of institutional endorsement. However, the modest post-workshop mean SI score of 5.05 is explained by the qualitative theme, "Navigating an Ambiguous Professional Landscape," where PSTs reported receiving conflicting messages from different university departments, lacking meaningful hands-on practice in their teacher education program, and a lack of clear AI policies in K-12 settings. This underscores the need for cohesive messaging among educators, especially through the implementation of clear and transparent policy in higher education and K-12 school districts (Cacho, 2024 ; Song, 2024 ). Similarly, while the workshop acted as a supportive resource for the PSTs, all interviewees called for more comprehensive and practical GenAI training to be embedded within their curriculum and practicum, aligning with recommendations from prior research on PSTs’ perceptions of GenAI (Thararattanasuwan & Prachagool, 2024 ; Wang et al., 2024 ). One of the most significant changes was observed in BI, which implies that the workshop successfully strengthened PSTs’ motivation to incorporate GenAI tools into their teaching and learning practices. This significant shift in BI is consistent with previous studies on technology acceptance, which indicate that educators are more inclined to adopt digital tools when they receive targeted training on how to use them pedagogically (Daher et al., 2021 ). The high average score post-workshop (5.67 out of 7) signifies a strong overall inclination to utilize these tools. This enthusiasm was echoed by interview subjects, with all PSTs expressing their intent to leverage GenAI tools to enhance both their efficiency and teaching methods. However, despite the promising rise in BI, it does not ensure actual adoption, as factors like institutional policies and the availability of resources for GenAI may still affect the implementation process (Zhang & Hou, 2024 ). This implies that future initiatives should focus not only on fostering intention but also on offering ongoing support and strategies to address potential barriers to successful implementation. The PSTs showed considerable engagement with GenAI tools for educational use even prior to the workshop, as indicated by a high average UB score of 5.92 out of 7 in the pre-workshop survey. This is consistent with qualitative feedback from participants, who noted using GenAI for various educational purposes. Thus, the workshop functioned not as an introduction to GenAI but as an opportunity to enhance or refine their understanding of its pedagogical benefits, usage strategies, and associated risks. 7. Implications This study offers important theoretical and practical implications. Theoretically, it presents an extended UTAUT model that incorporates trust, perceived risk, and TPACK as critical factors in educators' adoption of GenAI, offering a more nuanced framework for future research in this area. Practically, the findings provide a clear directive for MTEs and teacher preparation programs to prepare the next generation of teachers. It is no longer sufficient to simply acknowledge GenAI's existence. MTEs must proactively design learning experiences that equip PSTs with the critical competencies to use these tools effectively and ethically in the mathematics classroom. The positive shifts after the TPACK-based workshop in this study suggest that MTEs should model a pedagogy of inquiry with GenAI. This involves moving beyond showcasing GenAI as a tool for generating problems or lesson plans and instead teaching PSTs a process of critical co-creation. For mathematics, this means explicitly engaging PSTs in activities where they test GenAI’s computational accuracy, critique its pedagogical approaches, and learn to refine its outputs to align with specific mathematical learning goals and standards. The implication is a pedagogical shift from teaching about GenAI to teaching PSTs how to utilize GenAI as a flawed but powerful partner. Furthermore, this study implies that MTEs should cultivate a professionally and ethically responsible disposition toward GenAI. MTEs should intentionally design activities that highlight GenAI's limitations and potential for error in mathematical contexts. By doing so, they prepare PSTs to act as expert critics and validators of AI-generated content. This includes fostering a deep understanding of ethical issues, such as how student dependency on AI for answers can undermine the development of mathematical reasoning and problem-solving skills, and how using these tools intersects with professional and ethical responsibilities like protecting student data. Finally, this research implies that MTEs must assume a leadership role in resolving the "ambiguous professional landscape" their students are navigating. This extends the MTE's responsibility beyond their own classroom. MTEs are uniquely positioned to collaborate with higher education administrations and local K-12 school districts to help shape emerging AI policies, ensuring they are pedagogically sound and supportive of innovative teaching. By acting as advocates and connectors, MTEs can help build the supportive and consistent institutional ecosystem that PSTs require to utilize GenAI for meaningful classroom practice, thereby shaping the future of mathematics education in the age of AI. 8. Limitations and Recommendations The findings of this study should be interpreted in light of several limitations. First, the small sample of 25 PSTs from a single Midwestern university, primarily focused on mathematics education, limits the generalizability of the results. While the statistically significant improvements are promising, they show evidence of the workshop's potential in a specific setting rather than a universally applicable outcome. Future research should replicate this study with larger, more diverse samples across multiple institutions and disciplines to test the broader applicability of this TPACK-based intervention and explore how different contexts shape GenAI adoption. Second, the study's design constrains the scope of the conclusions. The single-session, pre-post design captures only immediate perceptual shifts. Thus, the shifts in intentions to adopt GenAI should be understood as a measure of immediate motivation, not a predictor of long-term classroom practice. To address this, future longitudinal studies should be conducted to track how these initial positive perceptions translate into sustained teaching behaviors as PSTs enter their own classrooms. Furthermore, the study relied on self-reported data, meaning the significant increase in self-reported TPACK reflects a growth in PSTs' confidence rather than an objective measure of their competence. Therefore, future work should incorporate performance-based assessments, such as the expert evaluation of AI-generated lesson plans or observational data from classroom simulations, to triangulate self-reported gains with demonstrated skill. Finally, the data were collected from participants in three separate workshop sessions, introducing a potential clustering effect that could violate the assumption of independence in our statistical tests. Future research should use statistical methods that account for such group-level variations. Declarations This study received approval from Iowa State University’s Institutional Review Board (IRB) before any data were collected (approval no. 24-468). References Ajenaghughrure, I., Sousa, S., & Lamas, D. (2021). Psychophysiological Modeling of Trust In Technology. Proceedings of the ACM on Human-Computer Interaction , 5, 1–25. https://doi.org/10.1145/3459745 Al-zboon, H. S., Gasaymeh, A. M., & Al-Rsa'i, M. S. (2021). The attitudes of science and mathematics teachers toward the integration of information and communication technology (ICT) in their educational practice: the application of the Unified Theory of Acceptance and Use of Technology (UTAUT). World Journal of Education , 11(1), 75-85. https://doi.org/10.5430/wje.v11n1p75 Appavoo, P. (2020). Acceptance of technology in the classroom: A qualitative analysis of mathematics teachers’ perceptions. In Frontiers in Intelligent Computing: Theory and Applications: Proceedings of the 7th International Conference on FICTA (2018), Volume 2 (pp. 355-364). Springer Singapore. https://doi.org/10.1007/978-981-13-9920-6_37 Athaluri, S. A., Manthena, S. V., Kesapragada, V. K. M., Yarlagadda, V., Dave, T., & Duddumpudi, R. T. S. (2023). Exploring the boundaries of reality: investigating the phenomenon of artificial intelligence hallucination in scientific writing through ChatGPT references. Cureus , 15(4). https://doi.org/10.7759/cureus.37432 Aytekin, A., Özköse, H., & Ayaz, A. (2022). Unified theory of acceptance and use of technology (UTAUT) in mobile learning adoption : Systematic literature review and bibliometric analysis. COLLNET Journal of Scientometrics and Information Management, 16(1), 75–116. https://doi.org/10.1080/09737766.2021.2007037 Bae, H., Hur, J., Park, J., Choi, G. W., & Moon, J. (2024). Pre-service Teachers’ Dual Perspectives on Generative AI: Benefits, Challenges, and Integrating into Teaching and Learning. Online Learning , 28(3), Article 3. https://doi.org/10.24059/olj.v28i3.4543 Baidoo-Anu, D., & Ansah, L. O. (2023). Education in the era of generative artificial intelligence (AI): Understanding the potential benefits of ChatGPT in promoting teaching and learning. Journal of AI , 7(1), 52-62. https://dx.doi.org/10.2139/ssrn.4337484 Bazelais, P., Lemay, D. J., & Doleck, T. (2024). User acceptance and adoption dynamics of ChatGPT in educational settings. Eurasia Journal of Mathematics, Science and Technology Education , 20(2), em2393. https://doi.org/10.29333/ejmste/14151 Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology , 3, 101 - 77. https://doi.org/10.1191/1478088706qp063oa Cacho, R. (2024). Integrating Generative AI in University Teaching and Learning: A Model for Balanced Guidelines. Online Learning , 28(3), Article 3. https://doi.org/10.24059/olj.v28i3.4508 Chan, C. K. Y., & Hu, W. (2023). Students’ voices on generative AI: Perceptions, benefits, and challenges in higher education. International Journal of Educational Technology in Higher Education , 20(1), 43. https://doi.org/10.1186/s41239-023-00411-8 Choung, H., David, P., & Ross, A. (2022). Trust in AI and Its Role in the Acceptance of AI Technologies. International Journal of Human–Computer Interaction , 39, 1727 - 1739. https://doi.org/10.1080/10447318.2022.2050543 Creswell, J. W., & Plano Clark, V. L. (2018). Designing and conducting mixed methods research (3rd ed.). SAGE Publications. Creswell, J. W. (2021). A concise introduction to mixed methods research. SAGE Publications. Daher, W., Baya’a, N., & Anabousy, A. (2021). Emotions and self-efficacy as mediators of pre-service teachers’ adoption of digital tools. Emerging Science Journal , 5(5), 636-649. https://doi.org/10.28991/esj-2021-01301 Dalal, M., Archambault, L., & Shelton, C. (2017). Professional development for international teachers: Examining TPACK and technology integration decision making. Journal of Research on Technology in Education , 49(3–4), 117–133. https://doi.org/10.1080/15391523.2017.1314780 Dasari, D., Hendriyanto, A., Sahara, S., Suryadi, D., Muhaimin, L. H., Chao, T., & Fitriana, L. (2024, January). ChatGPT in didactical tetrahedron, does it make an exception? A case study in mathematics teaching and learning. In Frontiers in Education (Vol. 8, p. 1295413). Frontiers. https://doi.org/10.3389/feduc.2023.1295413 Deng, X., & Joshi, K. D. (2024). Promoting ethical use of generative AI in education. ACM SIGMIS Database: the DATABASE for Advances in Information Systems , 55(3), 6–11. https://doi.org/10.1145/3685235.3685237 Ellis, A. R., & Slade, E. (2023). A new era of learning: Considerations for ChatGPT as a tool to enhance statistics and data science education. Journal of Statistics and Data Science Education , 1-10. https://doi.org/10.1080/26939169.2023.2223609 Faruk, L. I. D., Rohan, R., Ninrutsirikun, U., & Pal, D. (2023, December). University students’ acceptance and usage of generative AI (ChatGPT) from a psycho-technical perspective. In Proceedings of the 13th International Conference on Advances in Information Technology (pp. 1-8). https://doi.org/10.1145/3628454.3629552 Figg, C. & Jaipal-Jamani, K. (2012). TPACK-in-Practice: Developing 21st century teacher knowledge. In P. Resta (Ed.), Proceedings of SITE 2012--Society for Information Technology & Teacher Education International Conference (pp. 4683-4689) . Austin, Texas, USA: Association for the Advancement of Computing in Education (AACE). Figg, C., & Jaipal-Jamani, K. (2013). Transforming classroom practice: Technology professional development that works! Teaching & Learning , 8(1), 87–98. https://doi.org/10.26522/tl.v8i1.431 Garland, R. (1991). The mid-point on a rating scale: Is it desirable? Marketing Bulletin , 2, 66-70. Gattupalli, S. S., Lee, W., Allessio, D., Crabtree, D., Arroyo, I., Woolf, B. P., & Woolf, B. (2023). Exploring pre-service teachers' perceptions of large language models-generated hints in online mathematics learning. In LLM @ AIED (pp. 151-162). https://ceur-ws.org/Vol-3487/paper10.pdf Getenet, S. (2024). Pre-service teachers and ChatGPT in multistrategy problem-solving: Implications for mathematics teaching in primary schools. International Electronic Journal of Mathematics Education , 19(1), em0766. https://doi.org/10.29333/iejme/14141 Gross, N. (2023). What ChatGPT tells us about gender: a cautionary tale about performativity and gender biases in AI. Social Sciences , 12(8), 435. https://doi.org/10.3390/socsci12080435 Guler, N. K., Dertli, Z. G., Boran, E., & Yildiz, B. (2024). An artificial intelligence application in mathematics education: Evaluating ChatGPT’s academic achievement in a mathematics exam. Pedagogical Research , 9(2). https://doi.org/10.29333/pr/14145 Gurer, M. D. (2021). Examining technology acceptance of pre-service mathematics teachers in Turkey: A structural equation modeling approach. Education and Information Technologies , 26(4), 4709-4729. https://doi.org/10.1007/s10639-021-10493-4 Gurer, M. D., & Akkaya, R. (2022). The influence of pedagogical beliefs on technology acceptance: a structural equation modeling study of pre-service mathematics teachers. Journal of Mathematics Teacher Education , 25(4), 479-495. https://doi.org/10.1007/s10857-021-09504-5 Hsu, H.-P., Mak, J., Werner, J., White-Taylor, J., Geiselhofer, M., Gorman, A., & Capurro, C. T. (2024). Preliminary study on pre-service teachers’ applications and perceptions of generative artificial intelligence for lesson planning. Journal of Technology and Teacher Education , 32(3), 409-437. https://www.learntechlib.org/primary/p/224413/ Jaipal-Jamani, K., & Figg, C. (2015). The framework of TPACK-in-practice: Designing content-centric technology professional learning contexts to develop teacher knowledge of technology-enhanced teaching (TPACK). Technological pedagogical content knowledge: Exploring, developing, and assessing TPCK , 137-163. Jaipal-Jamani, K., Figg, C., Collier, D., Gallagher, T., Winters, K. L., & Ciampa, K. (2018). Developing TPACK of university faculty through technology leadership roles. Italian Journal of Educational Technology , 26(1), 39-55. https://doi.org/10.17471/2499-4324/984 Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., ... & Kasneci, G. (2023). ChatGPT for good? On opportunities and challenges of large language models for education. Learning and Individual Differences , 103, 102274. https://doi.org/10.1016/j.lindif.2023.102274 Kale, U. (2018). Technology valued? Observation and review activities to enhance future teachers’ utility value toward technology integration. Computers & Education, 117 , 160-174. https://doi.org/10.1016/j.compedu.2017.10.007 Khan Academy. (2024). Annual report 2023-24. https://annualreport.khanacademy.org/ Kiyici, O. D., & Övez, F. D. (2021). Examination of technology acceptance and TPACK competencies of mathematics teachers who are involved in distance education practices during the pandemic process. Journal of Educational Technology and Online Learning , 4(4), 805-821. https://doi.org/10.31681/jetol.1012204 Lee, D., & Palmer, E. (2025). Prompt engineering in higher education: A systematic review to help inform curricula. International Journal of Educational Technology in Higher Education , 22(1), Article 1. https://doi.org/10.1186/s41239-025-00503-7 Li, C., Zhu, W., Xing, W., & Guo, R. (2024, March). Analyzing student attention and acceptance of conversational AI for math learning: Insights from a randomized controlled trial. In Proceedings of the 14th Learning Analytics and Knowledge Conference (pp. 836-842). https://doi.org/10.1145/3636555.3636895 Li, M. (2024). Integrating artificial intelligence in primary mathematics education: investigating internal and external influences on teacher adoption. International Journal of Science and Mathematics Education , 1-26. https://doi.org/10.1007/s10763-024-10515-w Li, P. H., Lee, H. Y., Cheng, Y. P., Starčič, A. I., & Huang, Y. M. (2023, August). Solving the self-regulated learning problem: Exploring the performance of ChatGPT in mathematics. In International Conference on Innovative Technologies and Learning (pp. 77-86). Cham: Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-40113-8_8 MacDowell, P., Moskalyk, K., Korchinski, K., & Morrison, D. (2024). Preparing Educators to Teach and Create With Generative Artificial Intelligence. Canadian Journal of Learning and Technology , 50 (4), 1–23. https://doi.org/10.21432/cjlt28606 Matzakos, N., Doukakis, S., & Moundridou, M. (2023). Learning mathematics with large language models: A comparative study with computer algebra systems and other tools. International Journal of Emerging Technologies in Learning (iJET) , 18(20), 51-71. https://doi.org/10.3991/ijet.v18i20.42979 Mayer, P., & Girwidz, R. (2019, July). Physics teachers' acceptance of multimedia applications—adaptation of the technology acceptance model to investigate the influence of TPACK on physics teachers' acceptance behavior of multimedia applications. In Frontiers in Education (Vol. 4, p. 73). Frontiers Media SA. https://doi.org/10.3389/feduc.2019.00073 Menon, D., & Shilpa, K. (2023). “Chatting with ChatGPT”: Analyzing the factors influencing users' intention to Use the Open AI's ChatGPT using the UTAUT model. Heliyon , 9(11). https://doi.org/10.1016/j.heliyon.2023.e20962 Meletiou-Mavrotheris, M., & Paparistodemou, E. (2024). Sustaining teacher professional learning in STEM: Lessons learned from an 18-year-long journey into TPACK-guided professional development. Education Sciences , 14(4), 402. https://doi.org/10.3390/educsci14040402 Mishra, P., & Koehler, M. J. (2006). Technological pedagogical content knowledge: A framework for teacher knowledge. Teachers College Record , 108(6), 1017-1054. https://doi.org/10.1111/j.1467-9620.2006.00684.x Mistretta, R. M. (2005). Integrating technology into the mathematics classroom: The role of teacher preparation programs. Mathematics Educator , 15(1), 18-24. Mukuka, A., & Alex, J. K. (2024). Profiling mathematics teacher educators’ readiness for digital technology integration: Evidence from Zambia. Journal of Mathematics Teacher Education , 28 (2), Article 2. https://doi.org/10.1007/s10857-024-09657-z Ndlovu, M., Ramdhany, V., Spangenberg, E. D., & Govender, R. (2020). Preservice teachers’ beliefs and intentions about integrating mathematics teaching and learning ICTs in their classrooms. ZDM , 52(7), Article 7. https://doi.org/10.1007/s11858-020-01186-2 Ning, Y., Zhang, C., Xu, B., Zhou, Y., & Wijaya, T. T. (2024). Teachers’ AI-TPACK: Exploring the relationship between knowledge elements. Sustainability , 16(3), 978. https://doi.org/10.3390/su16030978 Nyaaba, M., & Zhai, X. (2024). Generative AI professional development needs for teacher educators. Journal of AI , 8(1), 1–13. https://doi.org/10.61969/jai.1385915 Özbek, T., Wekerle, C., & Kollar, I. (2023). Fostering pre-service teachers’ technology acceptance–does the type of engagement with tool-related information matter?. Education and Information Technologies , 1-23. https://doi.org/10.1007/s10639-023-12047-2 Perienen, A. (2020). Frameworks for ICT integration in mathematics education-A teacher’s perspective. Eurasia Journal of Mathematics, Science and Technology Education , 16(6), em1845. https://doi.org/10.29333/ejmste/7803 Philemon, N. M., Chibisa, A., & Mabusela, M. S. (2022). Acceptance of the GeoGebra application in learning circle theorems. International Journal of Learning, Teaching and Educational Research , 21(12), 1-20. Preston, C. C., & Colman, A. M. (2000). optimal number of response categories in rating scales: Reliability, validity, discriminating power, and respondent preferences. Acta Psychologica , 104(1), 1-15. https://doi.org/10.1016/S0001-6918(99)00050-5 Rahman, M. M., & Watanobe, Y. (2023). ChatGPT for education and research: Opportunities, threats, and strategies. Applied Sciences , 13(9), 5783. https://doi.org/10.3390/app13095783 Razmerita, L. (2024). Human-AI Collaboration: A Student-centered perspective of generative AI use in higher education. In Proceedings of The 23rd European Conference on e-Learning , 23(1), Article 1. https://doi.org/10.34190/ecel.23.1.3008 Sánchez-Ruiz, L. M., Moll-López, S., Nuñez-Pérez, A., Moraño-Fernández, J. A., & Vega-Fleitas, E. (2023). ChatGPT challenges blended learning methodologies in engineering education: A case study in mathematics. Applied Sciences , 13(10), 6039. https://doi.org/10.3390/app13106039 Song, N. (2024). Higher education crisis: Academic misconduct with generative AI. Journal of Contingencies and Crisis Management , 32(1), e12532. https://doi.org/10.1111/1468-5973.12532 Song, Y., Kim, J., Liu, Z., Li, C., & Xing, W. (2025). Students’ perceived roles, opportunities, and challenges of a generative AI-powered teachable agent: A case of middle school math class. Journal of Research on Technology in Education , 0(0), 1–19. https://doi.org/10.1080/15391523.2024.2447727 Strzelecki, A. (2023). To use or not to use ChatGPT in higher education? A study of students’ acceptance and use of technology. Interactive Learning Environments , 1-14. https://doi.org/10.1080/10494820.2023.2209881 Tai, S.J. D. (2015). From TPACK-in-action workshops to classrooms: CALL competency developed and integrated. Language Learning & Technology , 19(1), 139–164. https://dx.doi.org/10125/44406 Thararattanasuwan, K., & Prachagool, V. (2024). Exploring perspectives of teacher students toward generative AI technologies. International Education Studies , 17(5), Article 5. https://doi.org/10.5539/ies.v17n5p22 Venkatesh, V., Brown, S. A., & Bala, H. (2013). Bridging the qualitative-quantitative divide: Guidelines for conducting mixed methods research in information systems. MIS Quarterly , 37(1), 21–54. https://www.jstor.org/stable/43825936 Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly , 27(3), 425–478. https://doi.org/10.2307/30036540 Wang, K., Ruan, Q., Zhang, X., Fu, C., & Duan, B. (2024). Pre-service teachers’ genAI anxiety, technology self-efficacy, and TPACK: Their structural relations with behavioral intention to design genAI-assisted teaching. Behavioral Sciences , 14(5), Article 5. https://doi.org/10.3390/bs14050373 Wardat, Y., Tashtoush, M. A., AlAli, R., & Jarrah, A. M. (2023). ChatGPT: A revolutionary tool for teaching and learning mathematics. Eurasia Journal of Mathematics, Science and Technology Education , 19(7), em2286. https://doi.org/10.29333/ejmste/13272 Weidinger, L., Rauh, M., Marchal, N., Manzini, A., Hendricks, L., Mateos-Garcia, J., Bergman, S., Kay, J., Griffin, C., Bariach, B., Gabriel, I., Rieser, V., & Isaac, W. (2023). Sociotechnical safety evaluation of generative AI systems. ArXiv , abs/2310.11986. https://doi.org/10.48550/arXiv.2310.11986 Wong, G. K. (2015). Understanding technology acceptance in pre-service teachers of primary mathematics in Hong Kong. Australasian Journal of Educational Technology , 31(6). https://doi.org/10.14742/ajet.1890 Xiong, Y., Shi, Y., Pu, Q., & Liu, N. (2023). More trust or more risk? User acceptance of artificial intelligence virtual assistant. Human Factors and Ergonomics in Manufacturing & Service Industries . https://doi.org/10.1002/hfm.21020 Yang, S., & Appleget, C. (2024). An exploration of preservice teachers’ perceptions of generative AI: Applying the technological acceptance model. Journal of Digital Learning in Teacher Education , 40(3), 159–172. https://doi.org/10.1080/21532974.2024.2367573 Yildiz, E., & Arpaci, I. (2024). Understanding pre-service mathematics teachers’ intentions to use GeoGebra: The role of technological pedagogical content knowledge. E ducation and Information Technologies , 1-22. https://doi.org/10.1007/s10639-024-12614-1 Yilmaz, F. G. K., Yilmaz, R., & Ceylan, M. (2024). Generative artificial intelligence acceptance scale: A validity and reliability study. International Journal of Human–Computer Interaction , 40(24), 8703-8715. https://doi.org/10.1080/10447318.2023.2288730 Yilmaz, R., & Yilmaz, F. G. K. (2023). Augmented intelligence in programming learning: Examining student views on the use of ChatGPT for programming learning. Computers in Human Behavior: Artificial Humans , 1(2), 100005. https://doi.org/10.1016/j.chbah.2023.100005 Zhang, W., & Hou, Z. (2024). College teachers’ behavioral intention to adopt artificial intelligence-assisted teaching systems. IEEE Access , 12, 152812–152824. https://doi.org/10.1109/ACCESS.2024.3445909 Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7622889","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":515476957,"identity":"78924b3c-0687-49a0-a8f6-6d65005db71e","order_by":0,"name":"Shristi Shrestha","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5ElEQVRIie3OsWrCUBTG8QOBZDmQ9QgleYUjF0I3XyVZnLoJwUFoIJBNXQt9ilLofEogmbo7OLS4OkQC4nApNTo4WK66Odz/di/fDw6AzXaPEYDEAOg7pz++hjD28v1DjuvL5DBiuZaEr9Mf+db6QdXOqm0mOgAv/yAT4WXNkhSMUelGJBUrwCo1ExqCJFlHIAJxOcnoKTKS8GVPYs2ocq9t5Jefs3BtJrDoiMvIDjJ9FhwDoZlwR5JCIZWY0tdM9Qscjh4vHOZsdjoY+PP6vRlvg9D3yreF8bCz3NvmNpvNZvuvP2h2Re453pCFAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0009-0007-7110-9651","institution":"Iowa State University","correspondingAuthor":true,"prefix":"","firstName":"Shristi","middleName":"","lastName":"Shrestha","suffix":""},{"id":515476958,"identity":"95583891-7dac-4eec-b7e0-aded3cb8592d","order_by":1,"name":"Jiyeong Yi","email":"","orcid":"https://orcid.org/0000-0002-3031-4786","institution":"Iowa State University","correspondingAuthor":false,"prefix":"","firstName":"Jiyeong","middleName":"","lastName":"Yi","suffix":""}],"badges":[],"createdAt":"2025-09-15 16:48:45","currentVersionCode":1,"declarations":{"humanSubjects":true,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":true,"humanSubjectConsent":true,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-7622889/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7622889/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":91522048,"identity":"4a657fea-9859-4b5a-9001-0b76b4e27984","added_by":"auto","created_at":"2025-09-17 10:32:27","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":460301,"visible":true,"origin":"","legend":"\u003cp\u003eUTAUT Model (Venkatesh et al., 2003)\u003c/p\u003e","description":"","filename":"Fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7622889/v1/5e28ca12bba6ae9c926ba78f.jpg"},{"id":91522045,"identity":"8a6b8599-7ab4-49e6-a3e6-d3b65d87311d","added_by":"auto","created_at":"2025-09-17 10:32:27","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":374980,"visible":true,"origin":"","legend":"\u003cp\u003eExtended UTAUT Model\u0026nbsp;\u003c/p\u003e","description":"","filename":"Fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7622889/v1/1d31eb266c42d218edc29f8d.jpg"},{"id":91523232,"identity":"5277f737-ae85-495d-b64d-026235e92203","added_by":"auto","created_at":"2025-09-17 10:40:27","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":294787,"visible":true,"origin":"","legend":"\u003cp\u003eResearch Design\u003c/p\u003e","description":"","filename":"Fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7622889/v1/b02e541cf91aecf7dbf1b514.jpg"},{"id":91522047,"identity":"1ce5f10c-8aed-46be-8de1-fa90ecf80b8a","added_by":"auto","created_at":"2025-09-17 10:32:27","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":925844,"visible":true,"origin":"","legend":"\u003cp\u003eScreenshots of Simulated Learning Activity with GenAI-powered Chatbot, Khanmigo\u003c/p\u003e","description":"","filename":"Fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7622889/v1/c340f3e7fc974a953590f12d.jpg"},{"id":91523233,"identity":"e53d5834-3552-4349-800d-b472f8f079a0","added_by":"auto","created_at":"2025-09-17 10:40:27","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":439660,"visible":true,"origin":"","legend":"\u003cp\u003eScreenshot of PST Individual Responses on Wooclap for a Pedagogical Discussion Prompt\u003c/p\u003e","description":"","filename":"Fig5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7622889/v1/7c802c44b8f61ec56d500bb0.jpeg"},{"id":91523235,"identity":"affd9eeb-bd55-40bc-a740-b7f6845d41bb","added_by":"auto","created_at":"2025-09-17 10:40:27","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":677150,"visible":true,"origin":"","legend":"\u003cp\u003eScreenshot of a PST Using ChatGPT to Generate Hints for a Word Problem\u003c/p\u003e","description":"","filename":"Fig6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7622889/v1/6b361495cf9fb51ec42c8a98.jpg"},{"id":91523426,"identity":"b4678ecd-259e-4d7c-8507-fe9286d91f53","added_by":"auto","created_at":"2025-09-17 10:48:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4206835,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7622889/v1/8d9813bc-c633-4631-a37c-a5731510af3e.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eTPACK-based Professional Development for the AI Era: Fostering Pre-service Teachers' Acceptance of Generative AI in Mathematics Classrooms\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe emergence of Generative Artificial Intelligence (GenAI) presents a significant pedagogical challenge for mathematics teacher educators (MTEs): how do we prepare pre-service teachers (PSTs) to critically and effectively use this emerging technology to support mathematical learning? The swift progress of GenAI is significantly transforming the educational landscape through the increasing integration of AI-powered resources into instructional practices. Research indicates that GenAI systems can help deliver differentiated instruction (Getenet, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and promote student-centered learning (Song et al., \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Educators are adopting GenAI chatbots, like ChatGPT, alongside GenAI tools specialized for teaching, like Magic School AI, to streamline instructional responsibilities and provide tailored learning experiences for students (Baidoo-Anu \u0026amp; Ansah, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Kasneci et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Yilmaz \u0026amp; Yilmaz, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In mathematics education, GenAI is recognized as a valuable resource for developing instructional content (Ellis \u0026amp; Slade, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), solving math problems (Guler et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and providing personalized feedback (Li et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003ePSTs have been found to be receptive to utilizing instructional materials created by GenAI, recognizing its benefits for planning lessons and designing assessments (Wang et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). However, these advantages do not come without concerns. PSTs have expressed apprehensions about GenAI\u0026rsquo;s reliability and its potential to undermine teacher autonomy (Yang \u0026amp; Appleget, \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). These voiced reservations from PSTs, coupled with findings that many teacher educators lack the competencies to facilitate AI literacy for PSTs (Nyaaba \u0026amp; Zhai, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), highlight a significant gap in teacher preparation. Thus, it is imperative to develop structured training frameworks that equip PSTs with the knowledge and skills needed to thoughtfully and effectively incorporate GenAI tools in their teaching practices (S\u0026aacute;nchez-Ruiz et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Wang et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eDespite rising adoption of GenAI in mathematics classrooms, there is a significant research gap regarding the motivations that drive teachers to embrace these tools for teaching mathematics. Most studies investigating the acceptance of GenAI technology within educational settings have primarily focused on student usage patterns (Strzelecki, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Faruk et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Yilmaz et al., \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In contrast, there has been limited research exploring how pre-service and in-service teachers perceive the integration of GenAI into their pedagogical activities (Wang et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Li, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Given teachers' crucial role in determining technology adoption in classrooms, understanding the perspectives, concerns, and competencies of PSTs regarding GenAI is vital for developing effective training programs.\u003c/p\u003e\u003cp\u003eEducators tend to find technology easier to use as they become more proficient (Mukuka \u0026amp; Alex, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). However, many teacher education programs lack hands-on opportunities for PSTs to develop and implement their GenAI knowledge in classroom settings (Macdowell et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). A lack of empirical research in the field mirrors this gap in teacher education programs. Although established technology acceptance frameworks like the Unified Theory of Acceptance and Use of Technology (UTAUT) model (Venkatesh et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2003\u003c/span\u003e) are frequently used to investigate technology adoption among educators (Gurer, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Perienen, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), there is a dearth of research applying these models to understand how professional development (PD) shapes PSTs' adoption of GenAI, particularly within mathematics classrooms. Prior work establishes that PD rooted in pedagogical frameworks like the Technological Pedagogical Content Knowledge (TPACK) model (Mishra \u0026amp; Koehler, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2006\u003c/span\u003e) effectively enhances teacher confidence and willingness to adopt new technologies (Daher et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Jaipal-Jamani \u0026amp; Figg, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). However, there is a scarcity of studies that systematically evaluate such learning programs for GenAI adoption.\u003c/p\u003e\u003cp\u003eThe UTAUT model serves as a robust foundation for studying technology adoption. However, it does not fully account for the unique challenges introduced by GenAI, namely user trust, perceived risk, and the pedagogical competencies and content knowledge required for effective classroom use. Therefore, this study employs an extended UTAUT framework that integrates the constructs of Trust, Perceived Risk, and TPACK. These additions are based on research identifying them as critical factors in the acceptance of AI-driven and educational technologies (Xiong et al., \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Yildiz \u0026amp; Arpaci, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This enhanced model provides a more holistic lens for analyzing how PSTs perceive the adoption of GenAI tools and the underlying factors influencing their adoption decisions.\u003c/p\u003e\u003cp\u003eThis study addresses the gaps in the literature by designing and examining a TPACK-based PD workshop focused on building PSTs' technological and pedagogical knowledge to integrate GenAI tools into mathematics instruction. Using a mixed-methods approach, it quantitatively assesses the program's impact on PSTs' acceptance of GenAI in mathematics education and qualitatively explores the factors shaping their perceptions. Specifically, the research questions that guide this study are:\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eHow does a TPACK-based PD workshop impact PSTs' perceptions of GenAI\u0026rsquo;s adoption in mathematics classrooms?\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eWhat factors influence PSTs\u0026rsquo; willingness to adopt GenAI tools in mathematics education contexts?\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003eThis study offers the research community an extended UTAUT framework, potentially providing a more nuanced lens for future GenAI adoption studies in educational contexts. For MTEs, the findings from the PD workshop may inform the design of pedagogically-grounded GenAI training programs. Ultimately, this work intends to inform strategies that enhance PSTs\u0026rsquo; GenAI competencies and support them in navigating the pedagogical and ethical challenges of GenAI adoption in mathematics instruction.\u003c/p\u003e"},{"header":"2. Theoretical Perspectives","content":"\u003cp\u003eThe successful integration of technology in education begins with its acceptance by users (Appavoo, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). While several models exist to explain technology acceptance, this study builds upon the UTAUT model. UTAUT, introduced by Venkatesh et al. (\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2003\u003c/span\u003e), is a comprehensive model that explains a user's intention to adopt a technology. The model posits that four core constructs directly influence user acceptance:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003ePerformance Expectancy (PE): The degree of belief that using the technology will enhance job performance.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eEffort Expectancy (EE): The perceived ease of using the technology.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eSocial Influence (SI): The social pressure an individual feels to use the technology.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eFacilitating Conditions (FC): The perceived availability of organizational and technical resources to support the technology's use.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eIn the UTAUT model, the first three constructs\u0026mdash;PE, EE, and SI\u0026mdash;are direct determinants of Behavioral Intention (BI), which is a user's intention to use the technology. This intention, along with FC, then predicts the actual Use Behavior (UB), i.e., the genuine adoption of the technology. The model also includes moderating variables such as age, gender, experience, and voluntariness of use that influence relationships between the constructs, as demonstrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eUTAUT has been validated in various education research studies (Aytekin et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Yee et al., 2021) and has been widely used to assess the acceptance of AI technologies like prediction and recommendation systems (Menon \u0026amp; Shilpa, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). However, the unique aspects of GenAI for educators require an extension of the UTAUT model. GenAI introduces significant risks, including data bias, misinformation, and privacy violations (Weidinger et al., \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), which undermine user trust (Menon \u0026amp; Shilpa, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Trust is crucial for enhancing user acceptance, while perceived risks can hinder it (Ajenaghughrure et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Choung et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Thus, their inclusion in UTAUT is critical for understanding GenAI adoption, an approach adopted and validated by Xiong and colleagues (\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Similarly, pedagogical competence is a critical determinant of technology adoption for educators. TPACK, introduced by Mishra and Koehler (\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), illustrates the vital interplay among technology, pedagogy, and content knowledge for effective technology integration in education. Research (Kiyici \u0026amp; \u0026Ouml;vez, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Mayer \u0026amp; Girwidz, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) has consistently established the significant influence of TPACK on PSTs\u0026rsquo; and mathematics teachers\u0026rsquo; technology acceptance and their perceptions of a tool\u0026rsquo;s usefulness and ease of use. Additionally, Wang et al. (\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) found that PSTs\u0026rsquo; self-perceived TPACK enhanced their acceptance of GenAI by improving their perceptions of its performance and effort expectancy. Therefore, this study employs an extended UTAUT framework that integrates these three key constructs\u0026mdash;Trust (TR), Perceived Risk (PR), and TPACK\u0026mdash;to create a more nuanced understanding of GenAI acceptance among PSTs, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. This extension deliberately omits the original model's moderating factors to maintain a sharper focus on the central constructs most relevant to the research questions.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"3. Literature Review","content":"\u003cp\u003eThis section synthesizes literature to establish the foundation for this research. It first explores the affordances and challenges of integrating GenAI in mathematics education, then examines key factors influencing technology acceptance in educational contexts, and concludes by reviewing the role of PD in enhancing technology adoption among PSTs.\u003c/p\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e3.1 GenAI in Mathematics Education\u003c/h2\u003e\u003cp\u003eGenAI\u0026rsquo;s integration into mathematics education offers significant potential but presents considerable challenges. GenAI supports educators in generating course materials and lesson plans (Bazelais et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Ellis \u0026amp; Slade, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). For students, it enhances learning by assisting in solving problems, explaining concepts, and providing personalized feedback and hints (Guler et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Dasari et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Li et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Song et al. (\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) found that GenAI-powered agents can foster student-centered learning by improving engagement and conceptual understanding. Gattupalli et al. (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) found that PSTs sometimes prefer GenAI\u0026rsquo;s step-by-step guidance over human-crafted hints in online math learning. Furthermore, GenAI can be used to differentiate instruction (Getenet, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), create engaging tasks (S\u0026aacute;nchez-Ruiz et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and complement traditional tools like Computer Algebra Systems (CAS) (Matzakos et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), with research highlighting its promise in teaching specific topics like geometry (Wardat et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and its positive reception among students (S\u0026aacute;nchez-Ruiz et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eHowever, these advantages are contested by significant ethical and pedagogical concerns that warrant careful consideration. Key challenges include risks to academic integrity (Rahman \u0026amp; Watanobe, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), the potential spread of misinformation due to hallucinations (Athaluri et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), the reinforcement of biases (Gross, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and the likelihood of exacerbating the digital divide (Chan \u0026amp; Hu, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In the context of mathematics education, particular limitations involve questionable computational accuracy when compared to specialized CAS tools (Matzakos et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), an overreliance on AI by students for answers, and pedagogical gaps such as the lack of adaptability to diverse learning styles (S\u0026aacute;nchez-Ruiz et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These issues indicate that GenAI should complement, rather than replace, traditional teaching methods (Dasari et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Ellis \u0026amp; Slade, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Thus, effective integration of GenAI in mathematics classrooms hinges on targeted training for teachers to navigate these limitations and leverage GenAI\u0026rsquo;s full potential.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Technology Acceptance in Educational Contexts\u003c/h2\u003e\u003cp\u003eTechnology acceptance explains how and why individuals embrace new technologies. The acceptance of technologies in education is a widely researched area, frequently analyzed through theoretical frameworks like the UTAUT model (Venkatesh et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2003\u003c/span\u003e) and its variants. This section presents literature to help understand the factors influencing the acceptance of GenAI and other technologies in general and math-specific educational settings.\u003c/p\u003e\u003cp\u003eGenAI acceptance research in higher education has shown that several key factors drive students' intention to use these tools. Across multiple studies, performance expectancy\u0026mdash;the perceived usefulness of technology\u0026mdash;consistently emerges as a primary determinant of adoption (Bazelais et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Faruk et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Strzelecki, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Wang et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Beyond utility, student acceptance is also influenced by habit (Strzelecki, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), hedonic motivation, i.e., enjoyment (Strzelecki, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and contextual factors like perceived humanness and novelty value (Faruk et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Similarly, Strzelecki (\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) identified facilitating conditions (e.g., technical support and resources) as a critical driver. From students\u0026rsquo; psychological standpoint, personality traits such as openness can positively affect usage, while neuroticism may act as a barrier (Faruk et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Researchers have also developed standardized instruments like the Generative Artificial Intelligence Acceptance Scale (GAIAS) (Yilmaz et al., \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) to better measure these constructs for educational contexts, validated with a sample of 627 university students.\u003c/p\u003e\u003cp\u003eWhile student adoption has been widely studied, research on educators remains less common. A notable study by Wang et al. (\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) examined PSTs\u0026rsquo; intention to use GenAI, finding that performance expectancy and social influence were the strongest predictors. Their extended UTAUT model also incorporated technology self-efficacy and TPACK, finding that higher levels in these areas increased confidence and reduced GenAI anxiety. Interestingly, some studies have found constructs like effort expectancy, i.e., the perceived ease of use, and facilitating conditions, to have no significant impact on behavioral intention to use GenAI in learning and teaching (Bazelais et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Wang et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This suggests that for some users, the perceived benefits of GenAI may outweigh the challenges of use or the need for support.\u003c/p\u003e\u003cp\u003eFor mathematics teachers and PSTs, perceived usefulness and performance expectancy consistently predict their intention to use technology (Al-zboon et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Gurer, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Philemon, 2022; Wong, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Effort expectancy was also found to be an important factor influencing adoption (Gurer, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Perinen, 2020; Philemon, 2022). Additionally, facilitating conditions and social influence are often cited as powerful influences on technology adoption (Wong, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Gurer, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Al-zboon et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). A study by Perienen (\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) noted that despite recognizing technology's pedagogical value for mathematics instruction, teachers' actual integration was minimal, with perceived ease of use and facilitating conditions being the primary factors impacting use. This highlights a persistent need for PD focused on pedagogical integration of technology in classrooms. Furthermore, factors such as teacher attitudes, pedagogical beliefs, and technology self-efficacy are crucial in predicting technology acceptance (Wong, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Gurer \u0026amp; Akkaya, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Li, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), as is the alignment of the technology with instructional goals, i.e., task-technology fit (Philemon et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). As GenAI tools enter mathematics education, early research shows positive perceptions among US college students (Li et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), but teacher adoption remains complex. A recent study on Chinese primary mathematics teachers by Li (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) found that positive teacher attitudes significantly impacted GenAI adoption.\u003c/p\u003e\u003cp\u003eWhile the literature provides a solid foundation for understanding technology acceptance in educational contexts, a significant gap persists at the intersection of GenAI technology, teachers, and mathematics education. Research on GenAI has predominantly centered on students, while studies in mathematics education have only begun exploring GenAI. Consequently, the factors influencing the acceptance and use of GenAI tools specifically among pre-service mathematics teachers remain largely unexplored.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Enhancing Technology Acceptance of PSTs with PD\u003c/h2\u003e\u003cp\u003eThe literature consistently emphasizes the crucial significance of training programs and PD in facilitating technology integration into instructional practices (Mistretta, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Ndlovu et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). While a plethora of studies examine the factors influencing technology acceptance among educators, there remains a noticeable gap in research focused on interventions designed to improve this acceptance (Kale, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn this vein, Daher and colleagues (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) conducted a noteworthy investigation into the effectiveness of targeted PD in enhancing technology acceptance. Their study evaluated how the preparation of PSTs for utilizing digital tools in mathematics and science instruction influenced their acceptance of these technologies. The preparation model centered on developing TPACK, emphasizing technical proficiency and pedagogical understanding of digital tools. The intervention program incorporated collaborative discussions, the creation of digital content, and reflective practices regarding implementing these tools in authentic classroom environments. Analysis using paired sample t-tests revealed significant increases in pre- and post-intervention scores across several dimensions, including perceived ease of use, perceived usefulness, attitudes, intent to use, actual usage, and self-efficacy. Notably, anxiety levels did not exhibit a significant change.\u003c/p\u003e\u003cp\u003eIn summary, the teacher preparation program effectively transformed PSTs' attitudes and perceptions, enhancing several aspects of technology acceptance regarding digital tools. Another study by \u0026Ouml;zbek and colleagues (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) also demonstrates that educational interventions focused on new technology positively impact their acceptance among PSTs. The researchers indicated that various forms of engagement with tool-specific information are crucial in promoting this acceptance. Specifically, their investigation revealed that both PST groups, one that engaged in hands-on learning tasks with a new digital tool and the other that read a blog post about it, experienced an increase in their intention to incorporate the technology into their lesson plans.\u003c/p\u003e\u003cp\u003eFurthermore, Yang and Appleget (\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) underscored the importance of structured integration of GenAI within teacher education programs. In their study, PSTs utilized GenAI tools to generate and assess read-aloud questions during a literacy methods course, applying their pedagogical and content knowledge in a practical learning setting. The findings indicated that 91% of PSTs recognized GenAI as a valuable instructional tool, appreciating its efficiency. However, the PST also expressed concerns regarding the possible adverse effects of the technology on teacher agency and creativity, as well as the questionable reliability of the technology. A positive correlation was identified between PSTs' engagement with GenAI and their intentions to incorporate it in future teaching, highlighting the critical need for AI literacy and pedagogical adaptation within teacher preparation curricula.\u003c/p\u003e\u003cp\u003eOverall, the relevant literature suggests that educational interventions are vital in enhancing PSTs' acceptance of technology. By integrating insights from the studies mentioned, we posit that professional learning opportunities that offer exposure to the application of GenAI in mathematics education, combined with interactive activities, could significantly augment PSTs' acceptance of this technology.\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Methods","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e4.1 Research Design\u003c/h2\u003e\u003cp\u003eThis study employed a sequential mixed-methods design (Creswell, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) to comprehensively understand PSTs\u0026rsquo; attitudes toward using GenAI in mathematics education. This approach has been known to be particularly effective in technology acceptance research in educational settings (Creswell \u0026amp; Clark, 2018; Venkatesh et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), as it integrates the strengths of quantitative data to measure outcomes with qualitative data to explore the underlying processes driving them. As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, the initial quantitative phase addressed Research Question 1 (RQ1) by assessing the effects of an intervention workshop on PSTs' perceptions. The subsequent qualitative phase addressed Research Question 2 (RQ2), exploring the underlying factors that shaped these perceptions. This sequential structure ensured that the quantitative data informed the qualitative data collection, particularly for designing the semi-structured interview guide.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Context and Participants\u003c/h2\u003e\u003cp\u003eThis study was conducted in two teacher education courses at a large public university in the Midwest United States. One course focused on mathematics methods, consisting of 12 students, and another was an educational technologies course with 15 students in one section and 14 students in another. The first author facilitated three 90-minute workshops, one in the mathematics methods course and two in each section of the learning technologies course. All 41 PSTs enrolled in these courses participated in the workshops as part of their regular coursework and were invited to participate in the research. A total of 28 PSTs (8 male, 20 female) consented to participate, forming a convenience sample. This sample included various academic levels: six sophomores, 12 juniors, eight seniors, and two master\u0026rsquo;s students. No significant demographic differences were observed between the 28 participants and the 13 non-participants, suggesting the study sample reasonably represented the students enrolled in the courses.\u003c/p\u003e\u003cp\u003eFollowing the workshop, all 28 participants were invited via email to volunteer for a follow-up interview to discuss their experiences in more depth. The first five PSTs who responded and aspired to serve as mathematics educators were selected for the interviews. This purposive sub-sample included three females and two males, representing a range of academic levels: one junior, three seniors, and one master's student. The group comprised four secondary mathematics PSTs and one elementary education PST pursuing a secondary mathematics endorsement. This selection provided a focused perspective from participants whose career paths aligned directly with the study's focus on mathematics education.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e4.3 TPACK-based PD Workshop\u003c/h2\u003e\u003cp\u003eThe intervention consisted of a 90-minute workshop to enhance PSTs\u0026rsquo; competence and willingness to integrate GenAI in mathematics instruction. The workshop's design was grounded in the TPACK-based Professional Learning Design Model (TPLDM), a framework developed by Figg and Jaipal-Jamani (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). The TPLDM was selected because it moves beyond traditional, technocentric PD by providing a structured approach that systematically integrates technological, pedagogical, and content knowledge. Research has demonstrated the model's effectiveness in building TPACK competencies among diverse educators, from university faculty to K-12 teachers (Jaipal-Jamani et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Tai, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). The 90-minute workshop guided the PSTs through a carefully structured sequence of interactive activities organized around the four key components of the TPLDM framework, described below.\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section3\"\u003e\u003ch2\u003e4.3.1 Modeling a Technology-Enhanced Activity Type\u003c/h2\u003e\u003cp\u003eThe workshop started by placing the PSTs in a scenario-based simulation where they assumed the role of a sixth-grade student tackling ratio practice problems on Khan Academy, an online educational platform. During this modeled activity, the PSTs worked through several ratio problems both with and without the support of Khanmigo, a GenAI chatbot integrated into Khan Academy. Khanmigo was chosen for this demonstration because of the widespread use and recognition of Khan Academy among school districts in the U.S., making it a relevant and relatable example for PSTs. During the 2023\u0026ndash;2024 school year, Khan Academy had nearly 975,800 licensed users across 577 U.S. school communities (Khan Academy, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe simulation employed a sequence of images that followed a narrative, allowing PSTs to observe and critically assess the tool\u0026rsquo;s ability to provide instant feedback, scaffold problem-solving strategies, and tailor instruction to meet student needs. For instance, during the activity demonstrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, some PSTs raised concerns that while Khanmigo effectively prompted the student to identify the correct numbers needed for a ratio problem (five turtles and eight total animals), it immediately provided the final answer. Instead of guiding the student to form the ratio themselves, the chatbot stated, \"So, the ratio of turtles to total animals is 5 for every 8.\"\u003c/p\u003e\u003cp\u003eAfter the simulation, a volunteer PST participated in a live, hands-on session with Khanmigo, mimicking a student's real-time navigation of the tool. The PSTs engaged with Khanmigo in various manners\u0026mdash;they used the chatbot's suggested prompts, communicated with it using language typical of a sixth grader (including slang and spelling mistakes), and exhibited behaviors resistant to learning. This interactive demonstration enabled PSTs to closely examine how real-time interactions with a GenAI tutor resemble the experiences of their perceived impression of a sixth-grade learner. Thus, this activity, which lasted approximately 25 minutes, contributed to PSTs' understanding of the complex relationship between GenAI tools (Khanmigo), pedagogical strategies (scaffolding and differentiation), and math content (ratios).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\u003ch2\u003e4.3.2 Integrating Pedagogical Dialogue\u003c/h2\u003e\u003cp\u003eAfter the modeling exercise, PSTs took part in timed discussions on instructional implications of GenAI, particularly examining the relationship among pedagogy, mathematical content, and GenAI technology. The discussions centered on two key questions: (1) How can Khanmigo facilitate differentiation for mixed-ability classrooms? (2) What could be some drawbacks to students engaging with GenAI bots like Khanmigo?\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAs illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, the PSTs initially answered these prompts on their own for approximately one minute using an interactive educational tool, Wooclap. This was followed by two-minute small group discussions, and then a whole-class conversation that took about two minutes to integrate various perspectives. This stage specifically focused on the pedagogical aspect of the TPACK framework, encouraging PSTs to reflect on how technology and teaching methods intersect and affect student learning in mathematics education.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section3\"\u003e\u003ch2\u003e4.3.3 Developing Technical Skills through Short Tool Demonstrations\u003c/h2\u003e\u003cp\u003eIn this part, PSTs were asked to individually respond, via Wooclap, to the prompt, \u0026ldquo;How are GenAI tools being used by teachers for mathematics instruction?\u0026rdquo; This reflective exercise aimed to elicit their perceptions and knowledge about using GenAI tools in mathematics teaching. Then, the first author showed live demonstrations for about 40 minutes to showcase the instructional uses of various GenAI tools: ChatGPT, Microsoft Copilot, Google Gemini, and Anthropic Claude Sonnet. The workshop illustrated specific instructional applications pertinent to mathematics education, which included:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eIdeating interactive story-based problems for specific learning standards and goals\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eAdjusting a middle school-level percentage problem about discounts to an elementary-level addition/subtraction problem\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eSimplifying a mathematical word problem to improve the question\u0026rsquo;s readability\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eGenerating hints to provide students with scaffolding to solve math problems\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eGenerating a graph that depicts the story of the race between the tortoise and the hare via piecewise linear functions\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eEach instructional application was showcased using at least two distinct GenAI tools (e.g., ChatGPT and Google Gemini), prompting comparisons of their outputs based on the same prompt. Through these demonstrations, the PSTs witnessed how GenAI tools could support lesson planning, differentiation, and instructional scaffolding, providing them with essential technical skills to implement these technologies in their future classrooms. Consequently, PSTs were provided fundamental technical skills, effectively merging technological and content knowledge crucial for real-world classroom applications.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section3\"\u003e\u003ch2\u003e4.3.4 Applying TPACK to Design Their Own Task\u003c/h2\u003e\u003cp\u003eThe workshop concluded with PSTs applying their integrated technological, pedagogical, and content knowledge by independently creating instructional materials using GenAI. The PSTs selected one or more GenAI tools that had been previously showcased to develop materials that were aligned with specific mathematics learning objectives. The tasks drew on earlier demonstrated use cases, which allowed the PSTs to transform the knowledge acquired during the workshop into tangible teaching resources.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e illustrates the submitted work of a participating PST during this exercise, where the individual employed ChatGPT to generate hints for a mathematical word problem. This design activity enabled the PSTs to synthesize their learning by applying GenAI to enhance pedagogical strategies within the context of mathematics, thereby enabling an advancement in their TPACK competencies.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e4.4 Data Collection and Analysis\u003c/h2\u003e\u003cp\u003ePre- and post-surveys were administered via Qualtrics immediately before and after the workshop. The survey instrument included 29 items (see Appendix) measuring the nine constructs of the extended UTAUT model (PE, EE, SI, FC, TR, PR, TPACK, BI, and UB). The instrument was created by adapting items from previously validated instruments in recent, relevant literature (Ning et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Xiong et al., \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Yildiz \u0026amp; Arpaci, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Yilmaz et al., 2023) to ensure validity of the measured constructs. The instrument used a 7-point Likert scale, instead of five, to improve the scale\u0026rsquo;s sensitivity (Preston \u0026amp; Coleman, 2000) and reduce bias (Garland, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1991\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eInitial descriptive statistics were performed to review pertinent variables. Specifically, mean, median, and standard deviation (SD) were calculated to assess central tendency and variability. Among the 28 participating PSTs, data from three were excluded due to incomplete survey responses, resulting in a final sample of 25 for inferential analysis. Composite scores for each model construct were computed by averaging their respective item scores, thus providing an overall assessment of each construct both pre- and post-workshop. The internal consistency and reliability of the adapted instrument were confirmed with the current study's data, with Cronbach\u0026rsquo;s alpha coefficients (α) for each construct, with a reliability threshold set at 0.7. Before conducting inferential analyses, the Shapiro-Wilk test assessed the normality of each composite score distribution. A paired-samples t-test (for normally distributed data) or a Wilcoxon signed-rank test (for non-normally distributed data) was used to compare pre- and post-survey scores to determine the intervention workshop's impact, with a significance threshold set at p\u0026thinsp;\u0026lt;\u0026thinsp;0.05.\u003c/p\u003e\u003cp\u003eQualitative data were gathered from two primary sources, i.e., semi-structured interviews and workshop responses, to provide a deep and contextualized understanding of PSTs\u0026rsquo; perceptions. Following the workshop, in-depth semi-structured interviews, lasting 40 minutes to an hour, were conducted with five PSTs specializing in mathematics education. The interviews were audio-recorded and transcribed verbatim. The interview protocol was aligned with the study\u0026rsquo;s theoretical framework to explore participants\u0026rsquo; attitudes towards adopting GenAI in educational contexts. Participants digitally submitted individual written responses to discussion prompts and images of hands-on tasks during the workshop via the Wooclap platform. For the three workshops\u0026rsquo; discussion prompts, 24 participating PSTs submitted their thoughts, collecting a total of 88 responses. The number of individual responses for each question was as follows: The first question, \"How can Khanmigo facilitate differentiation for mixed-ability classrooms?\" received 24 responses. The second question, \"What could be some drawbacks to students engaging with GenAI bots like Khanmigo?\" received 29 responses. Finally, the third question, \"How are GenAI tools being used by teachers for mathematics instruction?\" received 35 responses.\u003c/p\u003e\u003cp\u003eAn integrated thematic analysis (Braun \u0026amp; Clarke, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2006\u003c/span\u003e) was conducted to synthesize findings from all qualitative data sources. The analysis, managed in the Dedoose software, followed a multi-phase coding process to ensure a robust and comprehensive interpretation. An initial coding phase was conducted using a deductive approach, where the constructs of the extended UTAUT model served as a priori parent codes. This step ensured that the analysis aligned with the study\u0026rsquo;s theoretical framework. Then, a second inductive coding phase was conducted within and across the UTAUT constructs to identify emergent themes and patterns. This allowed for capturing nuanced perspectives not fully encompassed by the initial framework.\u003c/p\u003e\u003cp\u003eThe analysis was primarily conducted by the first author, so many strategies were employed to ensure the trustworthiness and rigor of the findings. The qualitative findings from the interviews and workshop responses were systematically compared with each other and against the quantitative survey data. This methodological triangulation was used to identify areas of convergence, where both data types told the same story, and expansion, where the qualitative data provided a deeper explanation for the quantitative results. This process created a more valid and holistic interpretation of the PSTs' perceptions. Peer debriefing served as the primary strategy for ensuring analytical rigor. The first author regularly met with the second author to review the codebook, discuss analytical decisions, and challenge emerging interpretations. This critical dialogue helped validate the thematic structure and minimize potential researcher bias. A detailed audit trail was maintained, documenting every step of the analytical process, from creating and revising the codebook to developing the final themes.\u003c/p\u003e\u003c/div\u003e"},{"header":"5. Results","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\u003ch2\u003e5.1 Shifts in PSTs\u0026rsquo; Technology Acceptance of GenAI in Mathematics Classrooms\u003c/h2\u003e\u003cp\u003eTo evaluate the impact of the PD workshop, shifts across eight constructs of the extended UTAUT model, except UB, were measured. The pre-workshop UB scores (M\u0026thinsp;=\u0026thinsp;5.92, SD\u0026thinsp;=\u0026thinsp;.81) captured a baseline of the PSTs\u0026rsquo; pre-existing use of GenAI tools. Since the pre- and post-surveys were administered immediately before and after the workshop, no change was anticipated in the use behavior of PSTs. So, inferential analysis was not performed on the UB scores. Internal consistency and reliability of the adapted instrument were confirmed with the study's data from pre- and post-workshop surveys, with Cronbach\u0026rsquo;s alpha coefficients for constructs showing acceptable (.70\u0026thinsp;\u0026le;\u0026thinsp;α\u0026thinsp;\u0026lt;\u0026thinsp;.80), good (.80\u0026thinsp;\u0026le;\u0026thinsp;α\u0026thinsp;\u0026lt;\u0026thinsp;.90), or excellent (α\u0026thinsp;\u0026ge;\u0026thinsp;.90) reliability, as shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cem\u003eShifts in Pre- and Post-workshop Survey Scores for Extended UTAUT Constructs\u003c/em\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"8\" nameend=\"c8\" namest=\"c1\"\u003e\u003cp\u003ePaired Samples t-test\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eConstruct\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003ePre-workshop\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003ePost-workshop\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eMean Score\u003c/p\u003e\u003cp\u003eDifference\u003c/p\u003e\u003cp\u003e(M)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eTest statistic (t)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003ep-value (p)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCronbach\u0026rsquo;s Alpha (α)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eMean (SD)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eCronbach\u0026rsquo;s Alpha (α)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMean (SD)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.89\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.33 (1.06)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.86\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5.70 (0.87)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.37\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-2.47\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.0212\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.80\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4.57 (1.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5.05 (0.87)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.48\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-2.41\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.0239\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.09 (1.13)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.95\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5.40 (1.09)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.31\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-2.35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.0272\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.96 (1.37)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.83\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.73 (1.40)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-0.23\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e1.80\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.0842\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTPACK\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.86\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.29 (0.99)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5.85 (0.79)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.56\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-2.99\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.0052\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.78\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.13 (0.90)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5.67 (0.80)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.55\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-4.67\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.0001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"8\" nameend=\"c8\" namest=\"c1\"\u003e\u003cp\u003eWilcoxon Signed-rank Test\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eConstruct\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003ePre-workshop\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003ePost-workshop\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eMedian\u003c/p\u003e\u003cp\u003eScore\u003c/p\u003e\u003cp\u003eDifference\u003c/p\u003e\u003cp\u003e(\u003cem\u003eM\u003c/em\u003e)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eTest Statistic (W)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003ep-value (p)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCronbach\u0026rsquo;s Alpha (α)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eMedian (SD)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eCronbach\u0026rsquo;s Alpha (α)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMedian (SD)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.85\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.33 (0.93)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.90\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e6.00 (0.73)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.67\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e9.50\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.0003\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.82\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4.67 (1.21)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.83\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5.67 (1.05)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1.00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e27.50\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.0199\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe descriptive and inferential statistics, summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, reveal a significant positive shift in PSTs\u0026rsquo; perceptions of GenAI following the workshop. Participants reported statistically significant improvements in all constructs measured, except PR (p\u0026thinsp;\u0026gt;\u0026thinsp;.05). The most substantial changes were observed in constructs directly related to the PSTs' intention to adopt GenAI (BI; M\u0026thinsp;=\u0026thinsp;.55, t = -4.67, p\u0026thinsp;=\u0026thinsp;.0001) and perception of available resources and support to use GenAI in educational contexts (FC; \u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.00, W\u0026thinsp;=\u0026thinsp;27.5, p\u0026thinsp;=\u0026thinsp;.0199). This was complemented by highly significant gains in PE, indicating that after the workshop, PSTs more strongly believed that GenAI would help achieve their academic and instructional goals (\u003cem\u003eM\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.67, W\u0026thinsp;=\u0026thinsp;9.5, p\u0026thinsp;=\u0026thinsp;.0003). Furthermore, participants\u0026rsquo; confidence in integrating GenAI, pedagogy, and math knowledge (TPACK) grew significantly (M\u0026thinsp;=\u0026thinsp;.56, t = -2.99, p\u0026thinsp;=\u0026thinsp;.0052).\u003c/p\u003e\u003cp\u003eSignificant improvements were also found in PSTs\u0026rsquo; perceived ease of use of GenAI (EE; M\u0026thinsp;=\u0026thinsp;0.37, t = -2.47, p\u0026thinsp;=\u0026thinsp;0.0212), perceived influence from important others to adopt GenAI in educational contexts (SI; M\u0026thinsp;=\u0026thinsp;.48, t = -2.41, p\u0026thinsp;=\u0026thinsp;.0239), and trust on the technology (TR; M\u0026thinsp;=\u0026thinsp;.31, t = -2.35, p\u0026thinsp;=\u0026thinsp;.0272). In contrast, the decrease in PR was not statistically significant (M = -0.23, t\u0026thinsp;=\u0026thinsp;1.80, p\u0026thinsp;=\u0026thinsp;.0842), suggesting that while other perceptions became more positive, PSTs\u0026rsquo; sense of risk associated with using GenAI did not decrease significantly.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003e5.2 Factors Influencing PSTs\u0026rsquo; Willingness to Adopt GenAI\u003c/h2\u003e\u003cp\u003eThe integrated thematic analysis of interview and workshop data revealed five overarching themes. These themes synthesize the constructs of the extended UTAUT model into a holistic narrative of the facilitators and barriers shaping the PSTs\u0026rsquo; perspectives. The themes are supported by excerpts from the interview and workshop, with participants referred to by pseudonyms.\u003c/p\u003e\u003cdiv id=\"Sec19\" class=\"Section3\"\u003e\u003ch2\u003e5.2.1 Theme 1: GenAI as a Collaborative Partner for Efficiency and Creativity\u003c/h2\u003e\u003cp\u003eThis was the most dominant theme, identified by all five interviewees and heavily supported by the workshop data, where 57 out of 88 submitted responses were related to GenAI's utility. This theme encapsulates PSTs' views on GenAI\u0026rsquo;s perceived usefulness (PE) and their current use behavior (UB), framing GenAI not just as a tool, but as a partner that enhances their professional practice.\u003c/p\u003e\u003cp\u003ePSTs consistently described using GenAI to offload the repetitive, time-consuming aspects of teaching. This focus on efficiency was detailed in the workshop. In response to a question about how they think math teachers are using GenAI, participants identified the rapid generation of instructional materials as the key benefit, specifically highlighting its ability to create problems (11 responses), lesson plans (7 responses), and differentiated materials (7 responses). This sentiment was perfectly captured during an interview with Liberty, a junior in elementary education. She explained how she planned to use these tools to maintain a healthy work-life balance in her future career. She saw GenAI as a practical assistant that would \"help me stay at school within my contracted hours.\"\u003c/p\u003e\u003cp\u003eBeyond simple efficiency, PSTs valued GenAI as a creative collaborator that could enhance the quality of their instruction. Caleb, a master's student, provided a vivid description of this dynamic process during his interview, sharing that he does not ask GenAI for a finished product but instead engages in a brainstorming session. He explained, \"You can use it almost like a whiteboard; you just throw an idea to it, and it will throw an idea back, and you just keep bouncing back and forth until you get what you want.\" This collaborative potential was also seen in its application for students, a point that emerged from the workshop discussion on differentiation. Participants envisioned tools like Khanmigo acting as an intelligent, personalized tutor capable of \"creating different tiers of problems for different tiers of learners\" and providing instant scaffolding that \"meets [students] where they are at with their knowledge.\" By offering \"guided help\" for struggling learners and novel challenges for advanced ones, PSTs saw GenAI as a direct partner in fostering a more creative and differentiated learning environment.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section3\"\u003e\u003ch2\u003e5.2.2 Theme 2: Navigating a Learning Curve with Cautious Engagement\u003c/h2\u003e\u003cp\u003eThis theme, central to the experience of all five interviewees, combines PSTs\u0026rsquo; perceived ease of use (EE) and trust in the GenAI technology (TR) and their developing TPACK. It captures the understanding that while GenAI is accessible, leveraging it effectively is a learned skill that requires critical oversight.\u003c/p\u003e\u003cp\u003eCaleb shared that the quality of the output is the user's responsibility, not the tool's. He explained, \"GenAI is very dumb. It's as smart as you are. If you give it a well-detailed design prompt, it will give you a well-detailed answer. But if you give it a generic prompt, it'll give you a generic response.\" This sentiment was echoed by Amanda, a senior, who noted that getting a useful output \"takes some practice\" and that \"you have to learn how to feed it the information so that you get what you want.\" This reframes ease of use not as a static property of the tool, but as a skill the user develops over time.\u003c/p\u003e\u003cp\u003eThis cautious engagement is rooted in a situational trust in the technology. Liberty described a practical workflow that perfectly illustrates this balance. While she doesn\u0026rsquo;t \"trust it 100%,\" she confidently uses it for creative tasks like generating story problems. She then applies her own expertise, explaining that \"if it doesn't give me exactly what I'm looking for, I can use what it gave me. I switch it up and do what I'm looking for.\" This sentiment was shared by Jade, another senior, who trusts GenAI to explain a concept but has \"trust issues with it\" when asking it to solve a specific math problem. These examples show PSTs are already developing a nuanced TPACK, learning to discern which tasks are appropriate for AI and which require their own critical evaluation and expert verification.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec21\" class=\"Section3\"\u003e\u003ch2\u003e5.2.3 Theme 3: Managing Risks of Inaccuracy and Dependency\u003c/h2\u003e\u003cp\u003eA significant concern, raised by all five interviewees and dominating the workshop discussion on drawbacks of Khanmigo (29 responses), was the inherent risk associated with GenAI (PR). This theme combines perceived risk of GenAI technology with skepticism around its usefulness, highlighting the ethical and pedagogical dilemmas PSTs foresee.\u003c/p\u003e\u003cp\u003eThe most frequently cited risk, mentioned in 22 of the 29 workshop responses and by all five interviewees, was the potential for student over-reliance leading to reduced cognitive engagement. PSTs expressed concerns that when students repeatedly turn to AI for instant answers instead of struggling with problems independently, they fail to engage in the deeper cognitive processing necessary for true understanding. Workshop participants voiced a fear that students would learn to manipulate the AI (11 responses), noting they might \"coax out an answer [from AI] without them trying the problem\" or use it for \"just giving the answer instead of giving an explanation as to how they can solve a specific problem.\" This superficial approach, as Serenity, a workshop participant, stated, leads to students \u0026ldquo;getting the answer without showing interest in learning the process.\" Compounding this issue was the risk of students learning from inaccurate AI-generated responses. Caleb, during his interview, elaborated on this with a powerful cognitive example, warning that if a student learns a concept incorrectly from an AI that confidently gives a wrong answer, \"it's going to be very hard to get rid of it,\" underscoring the high stakes of misinformation in foundational learning.\u003c/p\u003e\u003cp\u003eBeyond these pedagogical concerns, all interviewed PSTs identified concerns around data protection and student privacy. Amanda articulated this particularly well, worrying about the unseen data implications of using these tools with students. She explained, \"You have to be careful what you ask them to do on their school computers. You don't know what pulls information. With FERPA and the fact that they're minors, that's always like a concern because, me as an adult, it's that fine line where I still need to protect their data.\" This demonstrates a sophisticated awareness of professional responsibilities and the potential ramifications of uncritical GenAI adoption.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec22\" class=\"Section3\"\u003e\u003ch2\u003e5.2.4 Theme 4: Navigating an Ambiguous Professional Landscape\u003c/h2\u003e\u003cp\u003eThis theme, a major point of frustration for all five interviewees, encompasses the elements of influence PSTs receive from important others (SI) and their perception of available resources and support (FC). It highlights how PSTs are navigating a simultaneously encouraging and unsupportive professional landscape, receiving conflicting advice and lacking clear institutional guidance on how to proceed with GenAI.\u003c/p\u003e\u003cp\u003ePSTs are receiving explicit advocacy from within their teacher education programs. They acknowledged that some courses have made efforts to introduce GenAI through class activities like a \"Stump the AI\" exercise, where AI chatbots are challenged with questions they are likely to get wrong. Two PSTs shared experiences of assignments that required them to use GenAI to create lesson plans and then critique the AI-generated outputs. However, four out of the five interviewees described this exposure as superficial. Logan, a senior, expressed a desire for more structured and practical guidance: \"They told us that it exists, but they should have been teaching us how to use it to create more innovative lesson plans...so that it's not something that we have to figure out by ourselves later down the road when we're swamped with lesson planning.\"\u003c/p\u003e\u003cp\u003eThis gap in practical training is compounded by a confusing lack of consensus from the wider academic and professional communities. Amanda, a PST co-majoring in Psychology, vividly described this institutional inconsistency, noting how her professors' views were starkly divided by department. She explained that the Psychology department wants to \"pretend like it is not even a thing,\" while the Education department is \"trying to figure out how to use it.\" This creates a confusing environment where PSTs are simultaneously encouraged and discouraged from using the same technology. This ambiguity extends to the K-12 environment they are preparing to enter. Jade stated, \"I don't wanna be the only one who's using it [GenAI] or not using it and then finding out everyone else uses it or doesn't use it. Because it's something that a lot of people are very strong for, and then some don't want you using it at all. And just making sure I find that common level of what is expected of me from higher up people and districts.\" This highlights that a major barrier to adoption isn't just personal skill or belief, but the absence of a consistent professional framework for AI integration.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec23\" class=\"Section3\"\u003e\u003ch2\u003e5.2.5 Theme 5: A Commitment to Proactive and Pragmatic Integration\u003c/h2\u003e\u003cp\u003eThe final theme reflects the PSTs' ultimate stance, combining their pre-existing GenAI usage (UB), developing TPACK, and strong intentions to use the technology (BI). Voiced by all five interviewees, this theme shows how PSTs are not waiting passively for guidance but are instead proactively cultivating their own professional approach to using GenAI. PSTs are already integrating GenAI into their workflows as a strategic and pragmatic tool. Liberty provided an excellent example, explaining that she uses Magic School AI to \"quickly create presentations, lesson plans, and write professional emails.\" This proactive use is coupled with a self-aware development of their TPACK. PSTs were consciously building the skills to use these tools effectively while recognizing their limitations. Jade expressed this evolving confidence: \"Not fully confident right now, especially because I don't know of all the AI websites that are out there...I may have to just redo a whole bunch of things, such as learning what is fully reliable.\"\u003c/p\u003e\u003cp\u003eDuring the interview, Logan recounted his experience using GenAI the previous summer, when he co-taught a summer school program at an urban middle school as part of his university's outreach initiative. He stated, \"One of the lessons I taught this summer was a statistics unit, and I used ChatGPT and Copilot to design simple experiments for my students. I can see myself using these tools more for similar tasks in the future.\" This illustrates that PSTs are committed to developing their strategies for integrating GenAI.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"6. Discussion","content":"\u003cp\u003eThis study investigated the impact of a TPACK-based PD workshop on PSTs\u0026rsquo; acceptance of GenAI in mathematics classrooms. The quantitative findings demonstrate that the intervention led to statistically significant improvements across seven extended UTAUT constructs, except PR. This aligns with prior research indicating that TPACK-based PD effectively enhances technology acceptance (Daher et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eA primary outcome of the workshop was a significant boost in PSTs\u0026rsquo; perceived utility and capability to use GenAI. The highly significant increase in PE suggests the workshop illustrated GenAI\u0026rsquo;s practical benefits to the PSTs, a finding consistent with other intervention studies (Daher et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; \u0026Ouml;zbek et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This was reflected in the qualitative data, where PSTs framed GenAI as a \"collaborative partner\" for enhancing efficiency and creativity. These perceptions align with findings from Yang and Appleget (\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), where PSTs recognized GenAI as a useful teaching tool. However, this optimism did not come without criticism. PSTs remained skeptical about GenAI\u0026rsquo;s utility for complex mathematical reasoning and problem solving, corroborating research highlighting its computational limitations (Matzakos et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This enhanced sense of utility and a significant increase in self-reported TPACK created a strong foundation for adoption. The pre-workshop survey yielded a mean TPACK score of 5.29, indicating that many participants already felt confident in integrating GenAI with pedagogy and content before the intervention. The statistically significant increase in post-workshop scores suggests that the intervention expanded PSTs\u0026rsquo; knowledge, boosting their confidence. Direct engagement with GenAI tools and structured pedagogical discussions likely facilitated this growth, allowing PSTs to explore AI's potential in their teaching contexts. These results align with prior research indicating that TPACK-based PD enhances teachers\u0026rsquo; perceived ability to integrate technology into instruction (Dalal et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Meletiou-Mavrotheris \u0026amp; Paparistodemou, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). However, qualitative insights also revealed that some participants felt uncertain about implementing GenAI tools in real classrooms and expressed a need for more practical experience. This highlights the importance of continued exposure and structured practice to solidify PSTs\u0026rsquo; skills and confidence in leveraging GenAI, as Bae et al. (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) suggested. The significant increase in EE suggests that the workshop lowered barriers to entry for using the technology. The live demonstrations and hands-on tasks could have improved the PSTs\u0026rsquo; perceived ease of use. During the interviews, PSTs shared that moving beyond basic use of GenAI requires skills and practice to leverage the technology\u0026rsquo;s full potential. One of the skills noted was prompt engineering, validating calls to formally teach prompting strategies in higher education (Lee \u0026amp; Palmer, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eOne of the most insightful findings of this study is the dynamic between trust and risk. While PSTs\u0026rsquo; trust in GenAI (TR) for educational tasks significantly increased, their perception of its risks (PR) did not significantly decrease. This outcome can likely be attributed to the workshop's design, which facilitated a discussion of GenAI's risks without providing explicit strategies for their mitigation. By making potential threats more salient and concrete, the workshop did not alleviate fear but rather transformed it into a more informed, critical awareness. This is reflected in the modest post-workshop TR score (M\u0026thinsp;=\u0026thinsp;5.40) and the qualitative theme of \"cautious engagement,\" which indicates that PSTs trust GenAI for specific tasks (e.g., creative brainstorming) while remaining wary of its limitations (e.g., for high-stakes mathematical problem-solving). The interviewed PSTs emphasized the need for human oversight, a concept extensively advocated in the literature (Deng \u0026amp; Joshi, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Razmerita, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Ultimately, the PSTs\u0026rsquo; persistent risk perception reflects a sophisticated professional understanding of GenAI\u0026rsquo;s potential for misinformation and student dependency, a well-documented concern (Hsu et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Rahman \u0026amp; Watanobe, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This understanding is a more desirable outcome for teacher education than uncritical acceptance.\u003c/p\u003e\u003cp\u003eAdditionally, the findings highlight the critical role of the institutional context. The significant increases in SI and FC indicate that the workshop, which was delivered as a part of the PSTs\u0026rsquo; teacher education courses, was perceived as a form of institutional endorsement. However, the modest post-workshop mean SI score of 5.05 is explained by the qualitative theme, \"Navigating an Ambiguous Professional Landscape,\" where PSTs reported receiving conflicting messages from different university departments, lacking meaningful hands-on practice in their teacher education program, and a lack of clear AI policies in K-12 settings. This underscores the need for cohesive messaging among educators, especially through the implementation of clear and transparent policy in higher education and K-12 school districts (Cacho, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Song, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Similarly, while the workshop acted as a supportive resource for the PSTs, all interviewees called for more comprehensive and practical GenAI training to be embedded within their curriculum and practicum, aligning with recommendations from prior research on PSTs\u0026rsquo; perceptions of GenAI (Thararattanasuwan \u0026amp; Prachagool, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Wang et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eOne of the most significant changes was observed in BI, which implies that the workshop successfully strengthened PSTs\u0026rsquo; motivation to incorporate GenAI tools into their teaching and learning practices. This significant shift in BI is consistent with previous studies on technology acceptance, which indicate that educators are more inclined to adopt digital tools when they receive targeted training on how to use them pedagogically (Daher et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The high average score post-workshop (5.67 out of 7) signifies a strong overall inclination to utilize these tools. This enthusiasm was echoed by interview subjects, with all PSTs expressing their intent to leverage GenAI tools to enhance both their efficiency and teaching methods. However, despite the promising rise in BI, it does not ensure actual adoption, as factors like institutional policies and the availability of resources for GenAI may still affect the implementation process (Zhang \u0026amp; Hou, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This implies that future initiatives should focus not only on fostering intention but also on offering ongoing support and strategies to address potential barriers to successful implementation.\u003c/p\u003e\u003cp\u003eThe PSTs showed considerable engagement with GenAI tools for educational use even prior to the workshop, as indicated by a high average UB score of 5.92 out of 7 in the pre-workshop survey. This is consistent with qualitative feedback from participants, who noted using GenAI for various educational purposes. Thus, the workshop functioned not as an introduction to GenAI but as an opportunity to enhance or refine their understanding of its pedagogical benefits, usage strategies, and associated risks.\u003c/p\u003e"},{"header":"7. Implications","content":"\u003cp\u003eThis study offers important theoretical and practical implications. Theoretically, it presents an extended UTAUT model that incorporates trust, perceived risk, and TPACK as critical factors in educators' adoption of GenAI, offering a more nuanced framework for future research in this area. Practically, the findings provide a clear directive for MTEs and teacher preparation programs to prepare the next generation of teachers. It is no longer sufficient to simply acknowledge GenAI's existence. MTEs must proactively design learning experiences that equip PSTs with the critical competencies to use these tools effectively and ethically in the mathematics classroom. The positive shifts after the TPACK-based workshop in this study suggest that MTEs should model a pedagogy of inquiry with GenAI. This involves moving beyond showcasing GenAI as a tool for generating problems or lesson plans and instead teaching PSTs a process of critical co-creation. For mathematics, this means explicitly engaging PSTs in activities where they test GenAI\u0026rsquo;s computational accuracy, critique its pedagogical approaches, and learn to refine its outputs to align with specific mathematical learning goals and standards. The implication is a pedagogical shift from teaching about GenAI to teaching PSTs how to utilize GenAI as a flawed but powerful partner.\u003c/p\u003e\u003cp\u003eFurthermore, this study implies that MTEs should cultivate a professionally and ethically responsible disposition toward GenAI. MTEs should intentionally design activities that highlight GenAI's limitations and potential for error in mathematical contexts. By doing so, they prepare PSTs to act as expert critics and validators of AI-generated content. This includes fostering a deep understanding of ethical issues, such as how student dependency on AI for answers can undermine the development of mathematical reasoning and problem-solving skills, and how using these tools intersects with professional and ethical responsibilities like protecting student data.\u003c/p\u003e\u003cp\u003eFinally, this research implies that MTEs must assume a leadership role in resolving the \"ambiguous professional landscape\" their students are navigating. This extends the MTE's responsibility beyond their own classroom. MTEs are uniquely positioned to collaborate with higher education administrations and local K-12 school districts to help shape emerging AI policies, ensuring they are pedagogically sound and supportive of innovative teaching. By acting as advocates and connectors, MTEs can help build the supportive and consistent institutional ecosystem that PSTs require to utilize GenAI for meaningful classroom practice, thereby shaping the future of mathematics education in the age of AI.\u003c/p\u003e"},{"header":"8. Limitations and Recommendations","content":"\u003cp\u003eThe findings of this study should be interpreted in light of several limitations. First, the small sample of 25 PSTs from a single Midwestern university, primarily focused on mathematics education, limits the generalizability of the results. While the statistically significant improvements are promising, they show evidence of the workshop's potential in a specific setting rather than a universally applicable outcome. Future research should replicate this study with larger, more diverse samples across multiple institutions and disciplines to test the broader applicability of this TPACK-based intervention and explore how different contexts shape GenAI adoption. Second, the study's design constrains the scope of the conclusions. The single-session, pre-post design captures only immediate perceptual shifts. Thus, the shifts in intentions to adopt GenAI should be understood as a measure of immediate motivation, not a predictor of long-term classroom practice. To address this, future longitudinal studies should be conducted to track how these initial positive perceptions translate into sustained teaching behaviors as PSTs enter their own classrooms. Furthermore, the study relied on self-reported data, meaning the significant increase in self-reported TPACK reflects a growth in PSTs' confidence rather than an objective measure of their competence. Therefore, future work should incorporate performance-based assessments, such as the expert evaluation of AI-generated lesson plans or observational data from classroom simulations, to triangulate self-reported gains with demonstrated skill. Finally, the data were collected from participants in three separate workshop sessions, introducing a potential clustering effect that could violate the assumption of independence in our statistical tests. Future research should use statistical methods that account for such group-level variations.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cspan\u003eThis study received approval from Iowa State University\u0026rsquo;s Institutional Review Board (IRB) before any data were collected (approval no. 24-468).\u003c/span\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAjenaghughrure, I., Sousa, S., \u0026amp; Lamas, D. (2021). Psychophysiological Modeling of Trust In Technology. \u003cem\u003eProceedings of the ACM on Human-Computer Interaction\u003c/em\u003e, 5, 1\u0026ndash;25. https://doi.org/10.1145/3459745\u003c/li\u003e\n\u003cli\u003eAl-zboon, H. S., Gasaymeh, A. M., \u0026amp; Al-Rsa\u0026apos;i, M. S. (2021). The attitudes of science and mathematics teachers toward the integration of information and communication technology (ICT) in their educational practice: the application of the Unified Theory of Acceptance and Use of Technology (UTAUT). \u003cem\u003eWorld Journal of Education\u003c/em\u003e, 11(1), 75-85. https://doi.org/10.5430/wje.v11n1p75\u003c/li\u003e\n\u003cli\u003eAppavoo, P. (2020). Acceptance of technology in the classroom: A qualitative analysis of mathematics teachers\u0026rsquo; perceptions. In Frontiers in Intelligent Computing: Theory and Applications: \u003cem\u003eProceedings of the 7th International Conference on FICTA (2018), Volume 2 (pp. 355-364). \u003c/em\u003eSpringer Singapore. https://doi.org/10.1007/978-981-13-9920-6_37\u003c/li\u003e\n\u003cli\u003eAthaluri, S. A., Manthena, S. V., Kesapragada, V. K. M., Yarlagadda, V., Dave, T., \u0026amp; Duddumpudi, R. T. S. (2023). Exploring the boundaries of reality: investigating the phenomenon of artificial intelligence hallucination in scientific writing through ChatGPT references. \u003cem\u003eCureus\u003c/em\u003e, 15(4). https://doi.org/10.7759/cureus.37432\u003c/li\u003e\n\u003cli\u003eAytekin, A., \u0026Ouml;zk\u0026ouml;se, H., \u0026amp; Ayaz, A. (2022). Unified theory of acceptance and use of technology (UTAUT) in mobile learning adoption : Systematic literature review and bibliometric analysis. \u003cem\u003eCOLLNET Journal of Scientometrics and Information Management,\u003c/em\u003e 16(1), 75\u0026ndash;116. https://doi.org/10.1080/09737766.2021.2007037\u003c/li\u003e\n\u003cli\u003eBae, H., Hur, J., Park, J., Choi, G. W., \u0026amp; Moon, J. (2024). Pre-service Teachers\u0026rsquo; Dual Perspectives on Generative AI: Benefits, Challenges, and Integrating into Teaching and Learning. \u003cem\u003eOnline Learning\u003c/em\u003e, 28(3), Article 3. https://doi.org/10.24059/olj.v28i3.4543\u003c/li\u003e\n\u003cli\u003eBaidoo-Anu, D., \u0026amp; Ansah, L. O. (2023). Education in the era of generative artificial intelligence (AI): Understanding the potential benefits of ChatGPT in promoting teaching and learning. \u003cem\u003eJournal of AI\u003c/em\u003e, 7(1), 52-62. https://dx.doi.org/10.2139/ssrn.4337484\u003c/li\u003e\n\u003cli\u003eBazelais, P., Lemay, D. J., \u0026amp; Doleck, T. (2024). User acceptance and adoption dynamics of ChatGPT in educational settings. \u003cem\u003eEurasia Journal of Mathematics, Science and Technology Education\u003c/em\u003e, 20(2), em2393. https://doi.org/10.29333/ejmste/14151\u003c/li\u003e\n\u003cli\u003eBraun, V., \u0026amp; Clarke, V. (2006). Using thematic analysis in psychology. \u003cem\u003eQualitative Research in Psychology\u003c/em\u003e, 3, 101 - 77. https://doi.org/10.1191/1478088706qp063oa\u003c/li\u003e\n\u003cli\u003eCacho, R. (2024). Integrating Generative AI in University Teaching and Learning: A Model for Balanced Guidelines. \u003cem\u003eOnline Learning\u003c/em\u003e, 28(3), Article 3. https://doi.org/10.24059/olj.v28i3.4508\u003c/li\u003e\n\u003cli\u003eChan, C. K. Y., \u0026amp; Hu, W. (2023). Students\u0026rsquo; voices on generative AI: Perceptions, benefits, and challenges in higher education. \u003cem\u003eInternational Journal of Educational Technology in Higher Education\u003c/em\u003e, 20(1), 43. https://doi.org/10.1186/s41239-023-00411-8\u003c/li\u003e\n\u003cli\u003eChoung, H., David, P., \u0026amp; Ross, A. (2022). Trust in AI and Its Role in the Acceptance of AI Technologies. \u003cem\u003eInternational Journal of Human\u0026ndash;Computer Interaction\u003c/em\u003e, 39, 1727 - 1739. https://doi.org/10.1080/10447318.2022.2050543\u003c/li\u003e\n\u003cli\u003eCreswell, J. W., \u0026amp; Plano Clark, V. L. (2018). \u003cem\u003eDesigning and conducting mixed methods research\u003c/em\u003e (3rd ed.). SAGE Publications.\u003c/li\u003e\n\u003cli\u003eCreswell, J. W. (2021). \u003cem\u003eA concise introduction to mixed methods research.\u003c/em\u003e SAGE Publications.\u003c/li\u003e\n\u003cli\u003eDaher, W., Baya\u0026rsquo;a, N., \u0026amp; Anabousy, A. (2021). Emotions and self-efficacy as mediators of pre-service teachers\u0026rsquo; adoption of digital tools. \u003cem\u003eEmerging Science Journal\u003c/em\u003e, 5(5), 636-649. https://doi.org/10.28991/esj-2021-01301\u003c/li\u003e\n\u003cli\u003eDalal, M., Archambault, L., \u0026amp; Shelton, C. (2017). Professional development for international teachers: Examining TPACK and technology integration decision making. \u003cem\u003eJournal of Research on Technology in Education\u003c/em\u003e, 49(3\u0026ndash;4), 117\u0026ndash;133. https://doi.org/10.1080/15391523.2017.1314780\u003c/li\u003e\n\u003cli\u003eDasari, D., Hendriyanto, A., Sahara, S., Suryadi, D., Muhaimin, L. H., Chao, T., \u0026amp; Fitriana, L. (2024, January). ChatGPT in didactical tetrahedron, does it make an exception? A case study in mathematics teaching and learning.\u003cem\u003e In Frontiers in Education\u003c/em\u003e (Vol. 8, p. 1295413). Frontiers. https://doi.org/10.3389/feduc.2023.1295413\u003c/li\u003e\n\u003cli\u003eDeng, X., \u0026amp; Joshi, K. D. (2024). Promoting ethical use of generative AI in education. \u003cem\u003eACM SIGMIS Database: the DATABASE for Advances in Information Systems\u003c/em\u003e, 55(3), 6\u0026ndash;11. https://doi.org/10.1145/3685235.3685237\u003c/li\u003e\n\u003cli\u003eEllis, A. R., \u0026amp; Slade, E. (2023). A new era of learning: Considerations for ChatGPT as a tool to enhance statistics and data science education.\u003cem\u003e Journal of Statistics and Data Science Education\u003c/em\u003e, 1-10. https://doi.org/10.1080/26939169.2023.2223609\u003c/li\u003e\n\u003cli\u003eFaruk, L. I. D., Rohan, R., Ninrutsirikun, U., \u0026amp; Pal, D. (2023, December). University students\u0026rsquo; acceptance and usage of generative AI (ChatGPT) from a psycho-technical perspective. \u003cem\u003eIn Proceedings of the 13th International Conference on Advances in Information Technology\u003c/em\u003e (pp. 1-8). https://doi.org/10.1145/3628454.3629552\u003c/li\u003e\n\u003cli\u003eFigg, C. \u0026amp; Jaipal-Jamani, K. (2012). TPACK-in-Practice: Developing 21st century teacher knowledge. \u003cem\u003eIn P. Resta (Ed.), Proceedings of SITE 2012--Society for Information Technology \u0026amp; Teacher Education International Conference (pp. 4683-4689)\u003c/em\u003e. Austin, Texas, USA: Association for the Advancement of Computing in Education (AACE).\u003c/li\u003e\n\u003cli\u003eFigg, C., \u0026amp; Jaipal-Jamani, K. (2013). Transforming classroom practice: Technology professional development that works! \u003cem\u003eTeaching \u0026amp; Learning\u003c/em\u003e, 8(1), 87\u0026ndash;98. https://doi.org/10.26522/tl.v8i1.431\u003c/li\u003e\n\u003cli\u003eGarland, R. (1991). The mid-point on a rating scale: Is it desirable?\u003cem\u003e Marketing Bulletin\u003c/em\u003e, 2, 66-70.\u003c/li\u003e\n\u003cli\u003eGattupalli, S. S., Lee, W., Allessio, D., Crabtree, D., Arroyo, I., Woolf, B. P., \u0026amp; Woolf, B. (2023). Exploring pre-service teachers\u0026apos; perceptions of large language models-generated hints in online mathematics learning. In \u003cem\u003eLLM @ AIED\u003c/em\u003e (pp. 151-162). https://ceur-ws.org/Vol-3487/paper10.pdf\u003c/li\u003e\n\u003cli\u003eGetenet, S. (2024). Pre-service teachers and ChatGPT in multistrategy problem-solving: Implications for mathematics teaching in primary schools. \u003cem\u003eInternational Electronic Journal of Mathematics Education\u003c/em\u003e, 19(1), em0766. https://doi.org/10.29333/iejme/14141\u003c/li\u003e\n\u003cli\u003eGross, N. (2023). What ChatGPT tells us about gender: a cautionary tale about performativity and gender biases in AI. \u003cem\u003eSocial Sciences\u003c/em\u003e, 12(8), 435. https://doi.org/10.3390/socsci12080435\u003c/li\u003e\n\u003cli\u003eGuler, N. K., Dertli, Z. G., Boran, E., \u0026amp; Yildiz, B. (2024). An artificial intelligence application in mathematics education: Evaluating ChatGPT\u0026rsquo;s academic achievement in a mathematics exam. \u003cem\u003ePedagogical Research\u003c/em\u003e, 9(2). https://doi.org/10.29333/pr/14145\u003c/li\u003e\n\u003cli\u003eGurer, M. D. (2021). Examining technology acceptance of pre-service mathematics teachers in Turkey: A structural equation modeling approach. \u003cem\u003eEducation and Information Technologies\u003c/em\u003e, 26(4), 4709-4729. https://doi.org/10.1007/s10639-021-10493-4\u003c/li\u003e\n\u003cli\u003eGurer, M. D., \u0026amp; Akkaya, R. (2022). The influence of pedagogical beliefs on technology acceptance: a structural equation modeling study of pre-service mathematics teachers. \u003cem\u003eJournal of Mathematics Teacher Education\u003c/em\u003e, 25(4), 479-495. https://doi.org/10.1007/s10857-021-09504-5\u003c/li\u003e\n\u003cli\u003eHsu, H.-P., Mak, J., Werner, J., White-Taylor, J., Geiselhofer, M., Gorman, A., \u0026amp; Capurro, C. T. (2024). Preliminary study on pre-service teachers\u0026rsquo; applications and perceptions of generative artificial intelligence for lesson planning. \u003cem\u003eJournal of Technology and Teacher Education\u003c/em\u003e, 32(3), 409-437. https://www.learntechlib.org/primary/p/224413/\u003c/li\u003e\n\u003cli\u003eJaipal-Jamani, K., \u0026amp; Figg, C. (2015). The framework of TPACK-in-practice: Designing content-centric technology professional learning contexts to develop teacher knowledge of technology-enhanced teaching (TPACK). \u003cem\u003eTechnological pedagogical content knowledge: Exploring, developing, and assessing TPCK\u003c/em\u003e, 137-163.\u003c/li\u003e\n\u003cli\u003eJaipal-Jamani, K., Figg, C., Collier, D., Gallagher, T., Winters, K. L., \u0026amp; Ciampa, K. (2018). Developing TPACK of university faculty through technology leadership roles. \u003cem\u003eItalian Journal of Educational Technology\u003c/em\u003e, 26(1), 39-55. https://doi.org/10.17471/2499-4324/984\u003c/li\u003e\n\u003cli\u003eKasneci, E., Se\u0026szlig;ler, K., K\u0026uuml;chemann, S., Bannert, M., Dementieva, D., Fischer, F., ... \u0026amp; Kasneci, G. (2023). ChatGPT for good? On opportunities and challenges of large language models for education. \u003cem\u003eLearning and Individual Differences\u003c/em\u003e, 103, 102274. https://doi.org/10.1016/j.lindif.2023.102274\u003c/li\u003e\n\u003cli\u003eKale, U. (2018). Technology valued? Observation and review activities to enhance future teachers\u0026rsquo; utility value toward technology integration. \u003cem\u003eComputers \u0026amp; Education, 117\u003c/em\u003e, 160-174. https://doi.org/10.1016/j.compedu.2017.10.007\u003c/li\u003e\n\u003cli\u003eKhan Academy. (2024). Annual report 2023-24. https://annualreport.khanacademy.org/\u003c/li\u003e\n\u003cli\u003eKiyici, O. D., \u0026amp; \u0026Ouml;vez, F. D. (2021). Examination of technology acceptance and TPACK competencies of mathematics teachers who are involved in distance education practices during the pandemic process. \u003cem\u003eJournal of Educational Technology and Online Learning\u003c/em\u003e, 4(4), 805-821. https://doi.org/10.31681/jetol.1012204\u003c/li\u003e\n\u003cli\u003eLee, D., \u0026amp; Palmer, E. (2025). Prompt engineering in higher education: A systematic review to help inform curricula. \u003cem\u003eInternational Journal of Educational Technology in Higher Education\u003c/em\u003e, 22(1), Article 1. https://doi.org/10.1186/s41239-025-00503-7\u003c/li\u003e\n\u003cli\u003eLi, C., Zhu, W., Xing, W., \u0026amp; Guo, R. (2024, March). Analyzing student attention and acceptance of conversational AI for math learning: Insights from a randomized controlled trial. \u003cem\u003eIn Proceedings of the 14th Learning Analytics and Knowledge Conference \u003c/em\u003e(pp. 836-842). https://doi.org/10.1145/3636555.3636895\u003c/li\u003e\n\u003cli\u003eLi, M. (2024). Integrating artificial intelligence in primary mathematics education: investigating internal and external influences on teacher adoption. \u003cem\u003eInternational Journal of Science and Mathematics Education\u003c/em\u003e, 1-26. https://doi.org/10.1007/s10763-024-10515-w\u003c/li\u003e\n\u003cli\u003eLi, P. H., Lee, H. Y., Cheng, Y. P., Starčič, A. I., \u0026amp; Huang, Y. M. (2023, August). Solving the self-regulated learning problem: Exploring the performance of ChatGPT in mathematics. \u003cem\u003eIn International Conference on Innovative Technologies and Learning\u003c/em\u003e (pp. 77-86). Cham: Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-40113-8_8\u003c/li\u003e\n\u003cli\u003eMacDowell, P., Moskalyk, K., Korchinski, K., \u0026amp; Morrison, D. (2024). Preparing Educators to Teach and Create With Generative Artificial Intelligence. \u003cem\u003eCanadian Journal of Learning and Technology\u003c/em\u003e, \u003cem\u003e50\u003c/em\u003e(4), 1\u0026ndash;23. https://doi.org/10.21432/cjlt28606\u003c/li\u003e\n\u003cli\u003eMatzakos, N., Doukakis, S., \u0026amp; Moundridou, M. (2023). Learning mathematics with large language models: A comparative study with computer algebra systems and other tools. \u003cem\u003eInternational Journal of Emerging Technologies in Learning (iJET)\u003c/em\u003e, 18(20), 51-71. https://doi.org/10.3991/ijet.v18i20.42979\u003c/li\u003e\n\u003cli\u003eMayer, P., \u0026amp; Girwidz, R. (2019, July). Physics teachers\u0026apos; acceptance of multimedia applications\u0026mdash;adaptation of the technology acceptance model to investigate the influence of TPACK on physics teachers\u0026apos; acceptance behavior of multimedia applications. \u003cem\u003eIn Frontiers in Education (Vol. 4, p. 73).\u003c/em\u003e Frontiers Media SA. https://doi.org/10.3389/feduc.2019.00073\u003c/li\u003e\n\u003cli\u003eMenon, D., \u0026amp; Shilpa, K. (2023). \u0026ldquo;Chatting with ChatGPT\u0026rdquo;: Analyzing the factors influencing users\u0026apos; intention to Use the Open AI\u0026apos;s ChatGPT using the UTAUT model. \u003cem\u003eHeliyon\u003c/em\u003e, 9(11). https://doi.org/10.1016/j.heliyon.2023.e20962\u003c/li\u003e\n\u003cli\u003eMeletiou-Mavrotheris, M., \u0026amp; Paparistodemou, E. (2024). Sustaining teacher professional learning in STEM: Lessons learned from an 18-year-long journey into TPACK-guided professional development. \u003cem\u003eEducation Sciences\u003c/em\u003e, 14(4), 402. https://doi.org/10.3390/educsci14040402\u003c/li\u003e\n\u003cli\u003eMishra, P., \u0026amp; Koehler, M. J. (2006). Technological pedagogical content knowledge: A framework for teacher knowledge.\u003cem\u003e Teachers College Record\u003c/em\u003e, 108(6), 1017-1054. https://doi.org/10.1111/j.1467-9620.2006.00684.x\u003c/li\u003e\n\u003cli\u003eMistretta, R. M. (2005). Integrating technology into the mathematics classroom: The role of teacher preparation programs. \u003cem\u003eMathematics Educator\u003c/em\u003e, 15(1), 18-24.\u003c/li\u003e\n\u003cli\u003eMukuka, A., \u0026amp; Alex, J. K. (2024). Profiling mathematics teacher educators\u0026rsquo; readiness for digital technology integration: Evidence from Zambia. \u003cem\u003eJournal of Mathematics Teacher Education\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(2), Article 2. https://doi.org/10.1007/s10857-024-09657-z\u003c/li\u003e\n\u003cli\u003eNdlovu, M., Ramdhany, V., Spangenberg, E. D., \u0026amp; Govender, R. (2020). Preservice teachers\u0026rsquo; beliefs and intentions about integrating mathematics teaching and learning ICTs in their classrooms. \u003cem\u003eZDM\u003c/em\u003e, 52(7), Article 7. https://doi.org/10.1007/s11858-020-01186-2\u003c/li\u003e\n\u003cli\u003eNing, Y., Zhang, C., Xu, B., Zhou, Y., \u0026amp; Wijaya, T. T. (2024). Teachers\u0026rsquo; AI-TPACK: Exploring the relationship between knowledge elements. \u003cem\u003eSustainability\u003c/em\u003e, 16(3), 978. https://doi.org/10.3390/su16030978\u003c/li\u003e\n\u003cli\u003eNyaaba, M., \u0026amp; Zhai, X. (2024). Generative AI professional development needs for teacher educators. \u003cem\u003eJournal of AI\u003c/em\u003e, 8(1), 1\u0026ndash;13. https://doi.org/10.61969/jai.1385915\u003c/li\u003e\n\u003cli\u003e\u0026Ouml;zbek, T., Wekerle, C., \u0026amp; Kollar, I. (2023). Fostering pre-service teachers\u0026rsquo; technology acceptance\u0026ndash;does the type of engagement with tool-related information matter?. \u003cem\u003eEducation and Information Technologies\u003c/em\u003e, 1-23. https://doi.org/10.1007/s10639-023-12047-2\u003c/li\u003e\n\u003cli\u003ePerienen, A. (2020). Frameworks for ICT integration in mathematics education-A teacher\u0026rsquo;s perspective. \u003cem\u003eEurasia Journal of Mathematics, Science and Technology Education\u003c/em\u003e, 16(6), em1845. https://doi.org/10.29333/ejmste/7803\u003c/li\u003e\n\u003cli\u003ePhilemon, N. M., Chibisa, A., \u0026amp; Mabusela, M. S. (2022). Acceptance of the GeoGebra application in learning circle theorems. \u003cem\u003eInternational Journal of Learning, Teaching and Educational Research\u003c/em\u003e, 21(12), 1-20.\u003c/li\u003e\n\u003cli\u003ePreston, C. C., \u0026amp; Colman, A. M. (2000). optimal number of response categories in rating scales: Reliability, validity, discriminating power, and respondent preferences. \u003cem\u003eActa Psychologica\u003c/em\u003e, 104(1), 1-15. https://doi.org/10.1016/S0001-6918(99)00050-5\u003c/li\u003e\n\u003cli\u003eRahman, M. M., \u0026amp; Watanobe, Y. (2023). ChatGPT for education and research: Opportunities, threats, and strategies. \u003cem\u003eApplied Sciences\u003c/em\u003e, 13(9), 5783. https://doi.org/10.3390/app13095783\u003c/li\u003e\n\u003cli\u003eRazmerita, L. (2024). Human-AI Collaboration: A Student-centered perspective of generative AI use in higher education. \u003cem\u003eIn Proceedings of The 23rd European Conference on e-Learning\u003c/em\u003e, 23(1), Article 1. https://doi.org/10.34190/ecel.23.1.3008\u003c/li\u003e\n\u003cli\u003eS\u0026aacute;nchez-Ruiz, L. M., Moll-L\u0026oacute;pez, S., Nu\u0026ntilde;ez-P\u0026eacute;rez, A., Mora\u0026ntilde;o-Fern\u0026aacute;ndez, J. A., \u0026amp; Vega-Fleitas, E. (2023). ChatGPT challenges blended learning methodologies in engineering education: A case study in mathematics. \u003cem\u003eApplied Sciences\u003c/em\u003e, 13(10), 6039. https://doi.org/10.3390/app13106039\u003c/li\u003e\n\u003cli\u003eSong, N. (2024). Higher education crisis: Academic misconduct with generative AI. \u003cem\u003eJournal of Contingencies and Crisis Management\u003c/em\u003e, 32(1), e12532. https://doi.org/10.1111/1468-5973.12532\u003c/li\u003e\n\u003cli\u003eSong, Y., Kim, J., Liu, Z., Li, C., \u0026amp; Xing, W. (2025). Students\u0026rsquo; perceived roles, opportunities, and challenges of a generative AI-powered teachable agent: A case of middle school math class. \u003cem\u003eJournal of Research on Technology in Education\u003c/em\u003e, 0(0), 1\u0026ndash;19. https://doi.org/10.1080/15391523.2024.2447727\u003c/li\u003e\n\u003cli\u003eStrzelecki, A. (2023). To use or not to use ChatGPT in higher education? A study of students\u0026rsquo; acceptance and use of technology.\u003cem\u003e Interactive Learning Environments\u003c/em\u003e, 1-14. https://doi.org/10.1080/10494820.2023.2209881\u003c/li\u003e\n\u003cli\u003eTai, S.J. D. (2015). From TPACK-in-action workshops to classrooms: CALL competency developed and integrated. \u003cem\u003eLanguage Learning \u0026amp; Technology\u003c/em\u003e, 19(1), 139\u0026ndash;164. https://dx.doi.org/10125/44406\u003c/li\u003e\n\u003cli\u003eThararattanasuwan, K., \u0026amp; Prachagool, V. (2024). Exploring perspectives of teacher students toward generative AI technologies. \u003cem\u003eInternational Education Studies\u003c/em\u003e, 17(5), Article 5. https://doi.org/10.5539/ies.v17n5p22\u003c/li\u003e\n\u003cli\u003eVenkatesh, V., Brown, S. A., \u0026amp; Bala, H. (2013). Bridging the qualitative-quantitative divide: Guidelines for conducting mixed methods research in information systems. \u003cem\u003eMIS Quarterly\u003c/em\u003e, 37(1), 21\u0026ndash;54. https://www.jstor.org/stable/43825936\u003c/li\u003e\n\u003cli\u003eVenkatesh, V., Morris, M. G., Davis, G. B., \u0026amp; Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. \u003cem\u003eMIS Quarterly\u003c/em\u003e, 27(3), 425\u0026ndash;478. https://doi.org/10.2307/30036540\u003c/li\u003e\n\u003cli\u003eWang, K., Ruan, Q., Zhang, X., Fu, C., \u0026amp; Duan, B. (2024). Pre-service teachers\u0026rsquo; genAI anxiety, technology self-efficacy, and TPACK: Their structural relations with behavioral intention to design genAI-assisted teaching. \u003cem\u003eBehavioral Sciences\u003c/em\u003e, 14(5), Article 5. https://doi.org/10.3390/bs14050373\u003c/li\u003e\n\u003cli\u003eWardat, Y., Tashtoush, M. A., AlAli, R., \u0026amp; Jarrah, A. M. (2023). ChatGPT: A revolutionary tool for teaching and learning mathematics. \u003cem\u003eEurasia Journal of Mathematics, Science and Technology Education\u003c/em\u003e, 19(7), em2286. https://doi.org/10.29333/ejmste/13272\u003c/li\u003e\n\u003cli\u003eWeidinger, L., Rauh, M., Marchal, N., Manzini, A., Hendricks, L., Mateos-Garcia, J., Bergman, S., Kay, J., Griffin, C., Bariach, B., Gabriel, I., Rieser, V., \u0026amp; Isaac, W. (2023). Sociotechnical safety evaluation of generative AI systems. \u003cem\u003eArXiv\u003c/em\u003e, abs/2310.11986. https://doi.org/10.48550/arXiv.2310.11986\u003c/li\u003e\n\u003cli\u003eWong, G. K. (2015). Understanding technology acceptance in pre-service teachers of primary mathematics in Hong Kong. \u003cem\u003eAustralasian Journal of Educational Technology\u003c/em\u003e, 31(6). https://doi.org/10.14742/ajet.1890\u003c/li\u003e\n\u003cli\u003eXiong, Y., Shi, Y., Pu, Q., \u0026amp; Liu, N. (2023). More trust or more risk? User acceptance of artificial intelligence virtual assistant. \u003cem\u003eHuman Factors and Ergonomics in Manufacturing \u0026amp; Service Industries\u003c/em\u003e. https://doi.org/10.1002/hfm.21020\u003c/li\u003e\n\u003cli\u003eYang, S., \u0026amp; Appleget, C. (2024). An exploration of preservice teachers\u0026rsquo; perceptions of generative AI: Applying the technological acceptance model. \u003cem\u003eJournal of Digital Learning in Teacher Education\u003c/em\u003e, 40(3), 159\u0026ndash;172. https://doi.org/10.1080/21532974.2024.2367573\u003c/li\u003e\n\u003cli\u003eYildiz, E., \u0026amp; Arpaci, I. (2024). Understanding pre-service mathematics teachers\u0026rsquo; intentions to use GeoGebra: The role of technological pedagogical content knowledge. E\u003cem\u003education and Information Technologies\u003c/em\u003e, 1-22. https://doi.org/10.1007/s10639-024-12614-1\u003c/li\u003e\n\u003cli\u003eYilmaz, F. G. K., Yilmaz, R., \u0026amp; Ceylan, M. (2024). Generative artificial intelligence acceptance scale: A validity and reliability study. \u003cem\u003eInternational Journal of Human\u0026ndash;Computer Interaction\u003c/em\u003e, 40(24), 8703-8715. https://doi.org/10.1080/10447318.2023.2288730\u003c/li\u003e\n\u003cli\u003eYilmaz, R., \u0026amp; Yilmaz, F. G. K. (2023). Augmented intelligence in programming learning: Examining student views on the use of ChatGPT for programming learning. \u003cem\u003eComputers in Human Behavior: Artificial Humans\u003c/em\u003e, 1(2), 100005. https://doi.org/10.1016/j.chbah.2023.100005\u003c/li\u003e\n\u003cli\u003eZhang, W., \u0026amp; Hou, Z. (2024). College teachers\u0026rsquo; behavioral intention to adopt artificial intelligence-assisted teaching systems. \u003cem\u003eIEEE Access\u003c/em\u003e, 12, 152812\u0026ndash;152824. https://doi.org/10.1109/ACCESS.2024.3445909\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Iowa State University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Generative AI, Pre-service Teachers, Mathematics Education, Professional Development, TPACK, Technology Acceptance","lastPublishedDoi":"10.21203/rs.3.rs-7622889/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7622889/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAs Generative AI (GenAI) becomes more prevalent, the need to prepare pre-service teachers (PSTs) for its use is a critical challenge for mathematics teacher educators (MTEs). Yet, little is known about how to best foster PSTs\u0026rsquo; adoption and critical use of GenAI in mathematics classrooms. This study addresses this gap by evaluating the impact of a 90-minute professional development workshop, grounded in the Technological Pedagogical Content Knowledge (TPACK) framework, on PSTs\u0026rsquo; technology acceptance in mathematics education. A mixed-methods design was employed, using pre- and post-surveys based on an extended Unified Theory of Acceptance and Use of Technology (UTAUT) model for quantitative data and semi-structured interviews and workshop discussions for qualitative data. Quantitative analysis revealed statistically significant positive shifts in many aspects of technology acceptance, except for PSTs\u0026rsquo; perceived risks of the technology. Qualitative analysis identified key facilitators to adoption, such as GenAI's utility for instructional efficiency, alongside significant barriers, including the lack of clear institutional guidance. The findings demonstrate that TPACK-based professional development opportunities can enhance PSTs\u0026rsquo; responsible adoption of GenAI in mathematics education. This study provides actionable implications for MTEs on designing pedagogically grounded training that addresses GenAI's practical applications and ethical complexities in mathematics classrooms.\u003c/p\u003e","manuscriptTitle":"TPACK-based Professional Development for the AI Era: Fostering Pre-service Teachers' Acceptance of Generative AI in Mathematics Classrooms","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-17 10:24:23","doi":"10.21203/rs.3.rs-7622889/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"8752d721-8285-4ad9-9ee1-129153956e38","owner":[],"postedDate":"September 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":54869753,"name":"Educational Psychology"},{"id":54869754,"name":"Artificial Intelligence and Machine Learning"}],"tags":[],"updatedAt":"2025-09-17T10:24:23+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-17 10:24:23","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7622889","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7622889","identity":"rs-7622889","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}