Exploring Barriers to AI Course Assistant Adoption: A Mixed-Methods Study on Student Non-Utilization

Exploring Barriers to AI Course Assistant Adoption: A Mixed-Methods Study on Student Non-Utilization | 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 Exploring Barriers to AI Course Assistant Adoption: A Mixed-Methods Study on Student Non-Utilization George Hanshaw, Christopher Sullivan This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5867866/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 25 Jul, 2025 Read the published version in Discover Artificial Intelligence → Version 1 posted 14 You are reading this latest preprint version Abstract This research investigates the factors behind the non-utilization of the AI course assistant Spark among students at Los Angeles Pacific University (LAPU). Despite AI’s proven ability to enhance academic performance, motivation, and efficiency, a significant portion of students choose not to engage with this technology. Through a mixed-methods exploratory approach, key barriers to adoption are identified, including perceptions of unnecessity, lack of interest, and unfamiliarity with AI tools. External challenges, such as technical issues and a preference for traditional learning methods, are also examined. By combining sentiment analysis with thematic analysis of student survey responses, the findings offer a comprehensive understanding of the reasons for non-use. Targeted strategies—improving communication about AI’s benefits, providing training to build familiarity, and integrating AI more seamlessly into coursework—are recommended to increase adoption. Addressing these barriers is crucial, as non-utilization not only limits individual academic growth but also undermines efforts to equip students with essential skills for an AI-driven future. This research contributes to the growing literature on AI in education and provides actionable insights for educators and institutions seeking to maximize the impact of AI on learning outcomes. ai course assistants educational technology student engagement mixed-methods research technology adoption learning enhancement barriers to technology use sentiment analysis thematic analysis higher education Figures Figure 1 Figure 2 Figure 3 Figure 4 Why Students Did Not Use AI Course Assistants In recent years, the integration of artificial intelligence (AI) in educational settings has shown tremendous potential in enhancing student learning experiences. AI-powered tools, such as course assistants, have been designed to provide personalized learning, immediate feedback, and adaptable instructional support that aligns with the needs of modern learners. Research has demonstrated that AI can significantly improve student engagement, motivation, and overall academic performance by offering tailored educational experiences (Chen, Zou, Cheng, & Xie, 2022 ; Deng & Yu, 2023 ). Hanshaw et al. (2024) found that students who use AI course assistants designed to employ the Socratic method, encouraging inquiry without completing assignments for them, achieve significantly higher grade point averages compared to students who do not use these tools. Wu and Yu (2024) found that students who utilize AI chatbots within their courses achieve better grade outcomes as well. Understanding and addressing the barriers to adoption, these tools risk becoming underutilized resources, exacerbating inequities in educational outcomes, and leaving many students under-prepared for a rapidly evolving technological landscape. Despite these benefits, the actual utilization of such AI tools by students remains inconsistent, with a significant portion opting not to engage with these resources. This paradox between the potential benefits and the actual usage of AI in education underscores a significant gap in our understanding of how students utilize AI. Previous studies have indicated that students' perceptions of AI, their familiarity with such technologies, and their confidence in using them play a significant rolein their adoption of these tools (Al-Abdullatif, 2023 ; Grassini, 2023 ). This study focuses on the AI course assistant Spark used at Los Angeles Pacific University (LAPU), which was introduced to support students in their courses and other university-level resources such as the writing hub. Preliminary findings from a pilot study indicate that students who utilized Spark in their courses demonstrated higher academic performance, motivation, and efficacy compared to non-users (Hanshaw et. al, 2024). However, a substantial number of students reported not using the AI assistant, raising critical questions about the factors contributing to this non-use. Understanding these factors is vital for maximizing the potential of AI in education generally, and for the adoption of AI course assistants specifically. If AI tools such as course assitants are underutilized, their ability to enhance educational outcomes is severely compromised. Therefore, this study aims to explore the technological, psychological, and educational reasons behind students' non-utilization of Spark and to develop strategies that could increase its adoption, thereby enhancing the overall learning experience. By employing a mixed-methods exploratory approach, this research seeks to identify both the barriers to and the opportunities for increasing AI course assistant usage among students. The insights gained will not only contribute to the literature on AI in education, but will also provide practical recommendations for educators and institutions aiming to foster greater engagement with AI technologies (Ilieva et al., 2023 ; Labadze, Grigolia, & Machaidze, 2023 ). Literature Review AI is a driving force behind many technologies used daily throughout society, such as household gadgets, fitness monitors, self-driving cars, and social media platforms. In educational settings, AI is revolutionizing the student learning experience (Labadze et al., 2023; Ilieva et al., 2023; Luckin et al. 2016). Despite these benefits, some students in our study did not use the AI course assistant, Spark, in their online classrooms. This literature review explores the reasons behind this underutilization, focusing on issues of awareness and interest. For students in their college years, who have been raised in an era marked by swift technological progress, AI is not an unfamiliar concept. These students have been brought up with AI-enabled tools and platforms, ranging from tailored learning applications to social media algorithms. (Hyesun et al., 2023) Still, adoption of AI in higher education is just beginning, “Data…shows a dramatic rise in AI usage since 2020 with a 150% increase from the prior 2 years 2020–2019.” (Crompton & Burke, 2023), yet students are not always as eager to engage with and adopt AI technology in education as they do with the other “Smart” interfaces they interact with regularly. The incorporation of AI in the realm of education has its roots in the influential work of Vygotsky (1978), particularly his theory of the Zone of Proximal Development (ZPD). This theory suggests that effective learning takes place in a zone where learners are sufficiently challenged within their capabilities yet still require guidance and support. AI tailors educational activities to the learners’ readiness level, ensuring they are neither understimulated nor overwhelmed, thus optimizing the learning experience. AI also possesses the ability to track the progression of the learning process, adapting in real-time as the learner interacts with the system. This continuous feedback loop enables the system to modify the learning activities to align with the learner’s evolving progress and needs. In the context of online education, adaptive systems employ AI to cultivate self-regulated learning among students. This approach encompasses a spectrum of learning abilities, including goal setting, self-monitoring, self-instruction, and self-reinforcement (Ng et al., 2023). Moreover, the incorporation of AI in differentiated learning has transformed personalized education. AI now equips educators with insights into students’ learning strategies, backgrounds, progress, and academic interests. Significantly, AI can aid in bridging the educational divide caused by inequality issues, thus fostering accessibility for all learners, including those with special educational needs (Ng et al., 2023). The evolution of technology in education has undergone a transformative journey. The earliest “teaching machines” were designed to facilitate learning in a manner reminiscent of a private tutor. In the 1950s, a significant advancement was made with the development of Self Adaptive Keyboard Instructors (SAKI). These systems were designed to adapt to student performance but could not be adjusted to suit the needs of individuals. A significant milestone in educational technology was reached in 1970 when AI was first applied to these educational systems, leading to the creation of Intelligent Tutoring Systems (ITS) (Tonbuloglu, 2023). These systems represented a significant leap forward in the field of educational technology, laying the foundation for the sophisticated, AI-driven learning platforms we see today. Despite their potential, these systems did not gain widespread recognition due to cost and accessibility issues. Applications AI has brought about a significant shift in the educational sector. The use of AI in education can enable the development of more personalized curricula on a global scale, making it relatively economical not only in terms of monetary investment but also in the time and attention required from educators (Tonbuloglu, 2023). AI plays a critical role in personalized learning experiences based on each individual’s needs and learning styles by identifying patterns in students’ learning behaviors, preferences, and achievements. Additionally, AI-enabled virtual assistants and chatbots in an online classroom can provide instant support and guidance to students, answering questions and analyzing student performance. These abilities of AI course assistants to accomplish these tasks has been shown to have a positive impact on student learning (Essel et al., 2022). In addition to personalized learning experiences through AI algorithms, instructors can analyze learners’ data to predict student performance and provide additional support. By leveraging this data, AI has reliably become a tool to predict performance and identify those at risk of failure, allowing instructors to intervene and provide further assistance. This allows for instructors to identify areas for improvement and tailor their teaching strategies accordingly, adjusting learning content and leveraging the use of teaching assistants and chatbots to provide immediate feedback and personalized testing. The advent of big data and AI has significantly reshaped management and higher education practices and is currently applicable within the following scenarios (Jia & Tu, 2023): Assessment of Students and Schools : These methods employ academic analytics to deliver a more customized educational experience. Grading and Evaluation : Technologies such as image recognition, computer vision, and prediction systems have streamlined the grading and evaluation of papers and exams. Personalized Intelligent Teaching : Intelligent teaching systems and learning analytics are employed to adapt teaching methods to individual student needs. Smart Facilities : The incorporation of facial recognition, speech recognition, virtual labs, augmented reality (A/R), virtual reality (V/R), and hearing and sensing technologies. Online and Mobile Remote Education : The utilization of edge computing, virtual personalized assistants, and real-time analysis, thusmaking learning more accessible and flexible. Students & Teachers The use of AI capabilities has had a noticeable impact on students, not only in providing personalized learning experiences, but also in various aspects such as interactivity, real-time feedback, and customization and improvement of resource management in teaching and learning environments (Jia & Tu, 2023). These customized learning environments may better stimulate students’ motivation to learn, as they perceive the alignment between learning content and their individual needs. Additionally, the interactivity and real-time feedback facilitated by AI applications can offer personalized pacing and error correction, allowing students to experience their learning progress more immediately. Such immediate feedback mechanisms can enhance students’ self-efficacy and consequently boost their motivation to learn (Crompton & Burke, 2023). It also provides a unique ability to scale to model the student's learning process, determine the necessary information through performance analysis, make decisions, and provide interaction by responding to student questions and directing questions to them. (Tonbuloglu, 2023) Some of the features that make the use of AI in educational processes different from other technologies are its ability to model the student's learning process, to determine the information needed through performance analysis and to make decisions, and to provide interaction by responding to student questions and directing questions to them (Tonbuloglu, 2023). AI-driven systems can develop custom learning profiles for each student and customize their learning journeys and materials based on their needs, ability, preferred mode of learning, and experience (Ng et al., 2023). The affordances of AI can also ensure that the support is timely without waiting for a person to be available. Also, AI course assistants can consider aspects of students’ academic ability, preferences, and best strategies for support (Crompton & Burke, 2023). AI has become an important tool for educators. AI-powered tools have evolved to be more educator-centric, helping teachers identify effective teaching methods based on students' learning data. These tools also automate administrative tasks, create assessments, and handle grading and feedback (Ng et al., 2023). This saves teachers time and increases efficiency. AI can assist teachers by generating recurring questions, providing students with personalized assistance, and facilitating communication with peers. Through data analysis, teachers can adjust their teaching approaches and customize learning resources to meet students' needs (Ng et al., 2023). Teachers can adapt their methods to changing learning scenarios and objectives, whether in traditional classrooms or online platforms. AI can also improve the teaching and learning process by transforming instructional design, evaluation, and learning environments (Tonbuloglu, 2023). Trust and Usefulness A user's perception of trustfulness and usefulness play an important role in how people / students interact with AI. Trust is an essential human mechanism that helps us navigate vulnerability, uncertainty, complexity, and ambiguity, all of which make up risk. It is a psychological state that enables us to be vulnerable based on the optimistic expectations of an outcome or someone’s actions. In contrast to conventional technologies that operate based on user commands and predetermined rules, AI operates with a level of independence. The unpredictable nature of AI, often referred to as a "black-box," underscores the critical importance of trust as users confront the complexity and potential hazards of AI’s decision processes (Hyesun et al., 2023). Within the realm of technology, trust is built upon three foundational dimensions: benevolence/helpfulness, integrity/reliability, and capability/effectiveness. The Technology Acceptance Model (TAM) initially posited that the two primary drivers of a user's intention to continue using technology are its perceived usefulness and ease of use. According to subsequent findings, trust enhances the perceived usefulness, which in turn boosts the likelihood of continued use. Trust also cultivates positive attitudes, further influencing the intention to use the technology. Moreover, trust has been shown to increase the perceived usefulness, which leads to more positive attitudes and a higher intention to use the technology. (Hyesun et al., 2023) As technology continues to advance, AI has emerged as a key player, introducing new dimensions of trust that go beyond traditional technologies. For AI, trust can be looked at from two perspectives: trust in AI's human-like qualities (essentially, the personality of the technology), and trust in the operational aspects of AI (its capability, dependability, and safety). However, perceived usefulness extends beyond trust alone. Users assess usefulness in terms of how effectively AI can meet their needs, enhance productivity, and deliver tangible benefits that justify its integration into their daily routines. When users perceive AI as genuinely useful, they are more likely to overlook potential risks and invest the necessary time and effort to learn and adapt to the technology. These facets, both in AI's human-like qualities and its operational aspects, greatly affect individuals’ perceptions of smart technology's usefulness and appeal, thereby influencing their willingness to use it. Cognitive trust is influenced by factors such as transparency, dependability, and the nature of the task, whereas emotional trust is shaped by assigning human characteristics to AI. Of all the factors, perceived ease of use had the most substantial overall impact on usage, followed by perceived usefulness and trust. (Hyesun et al., 2023) Other barriers While the implementation of AI in administrative and educational settings aims to improve efficiency and automate processes, its direct impact on students' critical thinking abilities is uncertain. Critical thinking goes beyond quick information processing and requires skills for independent thought, thorough understanding and sound judgment. Although AI is proficient at analyzing large amounts of data, it primarily involves inputting information. On the other hand, critical thinking involves deep evaluation, reasoning, and critiquing of information, which depend not just on AI but also on learning, reflection, and practical experience (Jia & Tu, 2023). There are also remaining issues about how this change in social interaction, particularly with the increasing presence of AI, plays out in developing additional skill sets related to relationships and thinking. While AI can assist in various educational tasks, its role in fostering social skills and self-efficacy is more complex. Research has shown that having beliefs in one's own ability to achieve goals, known as general self-efficacy, plays a crucial role in mental well-being, physical health, and the ability to change behavior (Jia & Tu, 2023). Social interaction, traditionally a human-centered activity, is fundamental in the development of self-efficacy. Through engaging in various social environments and interactions, individuals can observe, experience, and practice behaviors, which in turn reinforce their beliefs in their capabilities. As AI increasingly mediates social interactions, the dynamics of how self-efficacy is developed may shift. While AI can provide feedback and simulate certain social scenarios, it may lack the depth of human interaction needed to fully foster self-efficacy. Bandura's (1995) research has highlighted the influence of personal beliefs in one's abilities within social and cultural contexts, shaping individuals' life paths. The feedback, modeling, and encouragement received from others in social settings significantly contribute to strengthening one's self-efficacy, and AI's ability to replicate this may be limited. Other studies have emphasized the significance of life and career skills such as problem-solving, emotional intelligence, judgment, service orientation, negotiation, cognitive flexibility, as well as communication and teamwork skills in the fourth industrial revolution. AI can support the development of some of these skills, particularly in problem-solving and cognitive flexibility, but the human elements of communication, emotional intelligence, and teamwork require direct human interaction and intervention. Developing critical thinking through academic reading, another area AI can support, is essential for meeting the higher-order thinking requirements of 21st-century students (Ng et al., 2023). Therefore, while AI has its place in education, it is clear that human skills and intervention remain vital. Several potential risks and conflicts, such as privacy concerns, changes in how power is structured, and excessive control over programming and data have been identified bystudents and teachers due to the potential of creating misunderstandings or misleading information. There are concerns that AI could provide unreliable recommendations, which may negatively impact students' performance, especially if teachers depend solely on AI-driven technologies to predict and assess students' learning outcomes (Ng et al., 2023). It's important to recognize that AI-driven platforms can sometimes misinterpret users and provide inaccurate suggestions. Therefore, student learning outcomes and social interactions should not rely solely on AI interpretation. Lastly, the design of AI driven tools may not be sufficiently human-centered (or even student-centered), which may cause discomfort for students. For example, features such ash as eye tracking or facial expression analysis may feel like surveillance to students. AI-based misunderstandings, misleading information, limitations, and hidden ethical issues have been noted by researchers and experts in the field (Ng et al., 2023).Therefore, AI-competency for teachers is essential to enhance students' AI-driven online learning, and teachers need to enhancetheir skills and knowledge through continuous professional development. Most notably, “The researchers found a great lack in pedagogical and ethical implications of implementing AI in HE and that there was a need for more educational perspectives on AI developments from educators conducting this work” (Crompton & Burke, 2023). Conclusion AI has the potential to revolutionize educational experiences by providing personalized learning, immediate feedback, and adaptable instructional support; however, its actual utilization among students remains inconsistent. This literature review has highlighted various factors influencing the adoption of AI tools in educational settings, including technological familiarity, trust, and perceived usefulness on the part of both teachers and students. Understanding these factors is crucial for developing strategies to enhance student engagement with AI technologies, ultimately maximizing their educational benefits. What is lacking in the current body of research is an exploration of the reasons students may not engage with AI course assistants placed within their classroom. Therefore, by focusing on the AI course assistant Spark at Los Angeles Pacific University, this study aims to delve deeper into the specific reasons behind its underutilization. Understanding these factors is crucial for developing strategies to enhance student engagement with AI technologies, ultimately maximizing their educational benefits. The insights gained from this research will contribute to the broader literature on AI in education and offer practical recommendations for educators and institutions seeking to foster greater adoption of AI tools. Research Question RQ1: What are the primary factors (e.g., technological, psychological, educational) that contribute to some students not utilizing the AI course assistant Spark available in their online classroom? RQ2: What strategies can be implemented to increase the utilization of the AI course assistant Spark among students, thereby enhancing their learning experience and academic performance? Purpose AI, as outlined by Kim et al. (2020), is increasingly being designed to teach, interact, and adapt to human teaching and learning methods, offering significant potential to enhance educational experiences. LAPU's pilot study on the use of the AI assistant Spark demonstrated that student-usersnot only outperformed non-users academically, but also showed higher levels of motivation and efficacy (Hanshaw et al., 2024). Given these clear benefits and the relatively straightforward adoption process, our research seeks to explore why some students choose not to use the AI assistant. Understanding these reasons will enable LAPU designers and faculty members to refine the design and promotion of AI course assistants, thus showcasing the value and benefits in delivering personalized and contextualized learning experiences. After identifying the barriers and reasons behind why some students did not use the AI course assistant, we will identify ways to better educate and connect with students to increase the utilization rate, enabling them to reap the benefits of using AI course assistants. By addressing and mitigating these barriers, we hope to foster greater acceptance and integration of AI course assistants and AI technology in general within educational settings. Method Study Design This study employs a mixed-methods exploratory approach to understand the factors influencing students' decisions not to use the AI course assistant Spark at LAPU. By integrating both qualitative and quantitative methods, this framework allows for a comprehensive examination of student perceptions and behaviors, providing nuanced insights into the barriers and potential strategies for increasing AI course assistant adoption. Rationale for Mixed-Methods Approach The sentiment analysis quantifies the emotional tone of student responses, providing an overview of the general attitudes towards the AI tool. This component helps to identify the prevalence of different sentiments (positive, neutral, or negative) within the student population. Thematic analysis rigorously investigates the qualitative data, pinpointing distinct themes and sub-themes that elucidate the factors contributing to disengagement. This method explores the intrinsic motivations, impediments, and situational elements that influence the students’ choices. Justification for Exploratory Design The study aims to explore and identify factors affecting the adoption of AI course assistants in an educational setting without prior hypotheses. This exploratory nature is essential for understanding the complex and multifaceted reasons students may not engage with available AI technologies. The integration of quantitative and qualitative data provides a more holistic understanding of the issue, ensuring that both statistical trends and personal narratives are considered in the analysis. Objectives of the Framework The objective of this framework is to identify perceived or real barriers towards the use of the AI course assistants as well as to develop strategies to help students overcome these barriers and enhance the utilization of the AI course assistant. Identifying barriers requires us to uncover the technological, psychological, and educational factors that contribute to the non-use of the AI course assistant. After identifying barriers, we then formulate strategies to enhance the utilization of AI tools, thereby improving student learning experiences and outcomes. Participants All participants were active LAPU undergraduate and graduate students. LAPU is a fully online, accredited university that caters primarily to adult learners. With a focus on flexibility, LAPU offers a variety of programs designed to meet the needs of working professionals, parents, and individuals balancing multiple responsibilities. The student body is diverse, with learners from various backgrounds and experiences, many of whom are returning to education after a significant time away from formal learning. A total of 883 End-Of-Course (EOC) surveys were completed and submitted by students during the Summer 1 term in 2024. Of these, approximately 68% of students self-reported not using Spark, the AI course assistant. This study specifically analyzed the responses from 602 EOC surveys where students indicated they did not use Spark. The EOC surveys are administered to students in every course, and participation is both optional and anonymous. It is important to note that the 602 responses do not necessarily represent 602 unique students, as some students may have completed more than one course and thus submitted multiple surveys. Each student can submit the survey only once per course. The overall survey response rate was 42.1%. The respondents included students in both undergraduate and graduate-level courses.. While the demographic composition of the respondents is believed to reflect the diverse student body at LAPU, specific demographic data was not collected due to the anonymous nature of the EOC survey. This lack of demographic data may limit the ability to generalize the findings to the entire student population. Procedure We analyzed the data using both a sentiment and thematic analysis. The purpose of using these two methods was to gain a deeper and more holistic understanding of the reasons why some students did not utilize the AI course assistant in their classroom. This dual methodology was designed to provide a robust and nuanced understanding of the data collected from open-ended survey responses. Sentiment Analysis Sentiment analysis was utilized to gauge the overall emotional tone of the students' responses regarding their non-use of the AI course assistant. This technique allowed us to quantify the attitudes expressed in the text, categorizing them into positive, negative, or neutral sentiments. Applying sentiment analysis allowed us to identify the general mood and attitudes of the students towards the AI course assistant. This provided an initial layer of understanding, highlighting the prevalent emotional reactions thatcould indicate broader patterns of feelings or attitudes towards the use of artificial intelligence within an online classroom. Thematic Analysis We also chose to include a thematic analysis to explore the responses from a qualitative perspective. The thematic analysis gives us specific reasons and contextual factors behind the students' responses. This qualitative method involved systematically identifying, analyzing, and reporting patterns (themes) within the data. Thematic analysis enabled us to uncover the underlying reasons for the students' non-usage of the AI course assistant by identifying recurring themes and sub-themes in their responses. This approach provided rich, detailed insights into the various factors influencing student behavior, including potential barriers, misconceptions, and areas for improvement in communications about the AI course assistant.. By combining sentiment analysis with thematic analysis, we were able to achieve a comprehensive understanding of the students' experiences and perspectives. Sentiment analysis offered a broad overview of the emotional landscape, while thematic analysis provided depth and context to the specific issues raised by the students. This dual approach ensured that our study captured both the quantitative and qualitative dimensions of the data, leading to a more informed and actionable set of findings. Data Collection The data for this study was obtained by adding two questions to the End-of-Course (EOC) survey. The added questions about the non-use of Spark were designed to identify if a student used Spark and gather free writing responses from students who did not use Spark. Question 40 was a multiple choice question which asked, “Did you use Spark, the AI Course Assistant, while participating in your course?”. If the respondent selected “no” to Question 40, Question 42 became available to them: “What is the primary reason(s) you did not use Spark within your course?”. Data Preparation The survey responses were compiled into an Excel spreadsheet. The relevant data for this analysis was located in Column B of the spreadsheet, corresponding to "Question 41.” Any responses that were missing or null were excluded from the analysis to ensure the accuracy and relevance of the sentiment analysis. Sentiment Analysis To gain a quantitative understanding of the emotional tone of the students' responses regarding their non-use of the AI course assistant "Spark," we conducted a sentiment analysis using the TextBlob library. This analysis assigned a polarity score to each response, ranging from -1 (negative sentiment) to 1 (positive sentiment), with 0 indicating neutral sentiment. Data Processing Prior to analysis, responses underwent preprocessing to remove punctuation, convert text to lowercase, and exclude common stop words. These words are typically articles (a, an, the…), prepositions (in, on, at…), conjunctions (and, or, but…) and pronouns (I, you, he. she…). Responses lacking content or deemed irrelevant were excluded to ensure a focused analysis. Statistical Analysis and Visualization Descriptive statistics summarized the sentiment distribution. A histogram was employed to show the frequency of sentiment categories, while a pie chart illustrated the proportion of positive, neutral, and negative sentiments. Additionally, a word cloud visualized frequently occurring terms to highlight common themes. Limitations and Biases The sentiment analysis using TextBlob is subject to certain limitations, including its inability to capture sarcasm and nuanced context. Misclassification is possible when responses contain ambiguous language. Furthermore, the voluntary nature of survey participation introduces potential selection bias, and the absence of demographic data limits the generalizability of findings. The reliance on self-reported data introduces the possibility of response bias, where participants may not accurately or honestly report their true experiences or feelings, influenced by factors such as memory recall issues or social desirability bias (Podsakoff et al., 2003). These limitations highlight the need for cautious interpretation of the results and suggest areas for future research to address these potential biases." Instruments, Tools, and Software The analysis was conducted using Python, with the Pandas, TextBlob, Matlotlib, and WordCloud libraries. These libraries were chosen for their robustness and ease of use in handling text data and generating meaningful visualizations. Thematic Analysis Data Organization Comments from column "Question 41" were systematically extracted and compiled into a list for detailed analysis. This step ensured that the data was ready for the coding process. Familiarization with Data To gain a comprehensive understanding of the content and context, the comments were read multiple times. This iterative reading facilitated the identification of initial patterns and emergent themes within the data. Generation of Initial Codes Each comment was assigned an initial code, which acted as a descriptive label summarizing the core reason provided by the student for not using the AI course assistant. These codes were preliminary and subject to further refinement. Theme Identification The initial codes were analyzed to identify broader themes that encapsulated the reasons for non-use. This involved grouping similar codes together to form overarching themes such as "Perception of Unnecessity," "Lack of Interest or Motivation," "External Issues," "Forgetfulness or Awareness Issues," "Lack of Familiarity with AI Tools," "Technical or Usability Issues," and "Preference for Traditional Methods." Theme Review and Refinement The identified themes underwent a rigorous review process to ensure they accurately represented the data. This involved evaluating the coherence and distinctiveness of each theme and making necessary adjustments to refine their definitions and boundaries. Re-categorization of Comments Comments were re-categorized according to the refined themes to maintain consistency and accuracy. Any comments that did not fit within the predefined themes were categorized under "Other" to capture additional nuances. Definition and Naming of Themes Each theme was clearly defined and appropriately named to reflect the underlying reasons for non-use. Representative comments were selected to illustrate the key points within each theme, providing a rich description of the data. Report Production Frequency Analysis The frequency of each theme was calculated to quantify the prevalence of different reasons among the students. This quantitative aspect complemented the qualitative analysis, offering a clearer picture of the dominant themes. Visualization A bar chart was generated to visualize the frequency distribution of the identified themes. This graphical representation helped highlight the most common reasons for not using the AI course assistant. By adhering to this structured methodology, the thematic analysis yielded a robust understanding of the reasons behind students' non-use of the AI course assistant, offering valuable insights for improving its adoption and effectiveness. Results Sentiment analysis The sentiment analysis of the survey responses to Question 41, “What is the primary reason(s) you did not use Spark within your course?” revealed the following distribution: Neutral Sentiment: The majority of the responses (72.3%) were neutral, with a sentiment score of 0.00. Examples of neutral responses include statements like "I had no reason to use it" and "I did not need to use it." Positive Sentiment: A smaller portion of the responses (27.7%) were positive, with sentiment scores greater than 0. Examples of positive responses include "Did not need to. I was able to do research on my own," which had a sentiment score of 0.55. Negative Sentiment: There were no responses with negative sentiment scores, indicating a lack of strong negative feelings towards the tool. Visualizations The histogram (see Figure 1) shows the distribution of sentiment scores across the responses. Most scores were clustered around 0, confirming the predominance of neutral responses. Positive sentiment scores were present but less frequent, while negative sentiment scores were absent. The pie chart in Figure 2 illustrates the proportion of responses in each sentiment category. The distribution showed that neutral responses constituted the majority (72.3%), followed by positive responses (27.7%), with no negative responses. The word cloud (see Figure 3) visualized the most frequently occurring words and phrases from the survey responses. Common words included "need," "reason," and "use," reflecting the main themes in the reasons provided by respondents. Interpretation The sentiment analysis results indicate that most respondents did not have strong feelings, either positive or negative, about not using the tool. The predominance of neutral sentiment suggests that their reasons were more practical or situational rather than based on strong emotional responses. Positive sentiments, though less frequent, showed confidence in alternative methods or tools. Respondents expressed satisfaction with their ability to perform tasks without the tool, indicating that their needs were met through other means. The absence of negative sentiments suggests that there is no widespread dissatisfaction with the tool itself. Instead, the reasons for not using the tool seem to stem from a perceived lack of necessity rather than any inherent issues with the tool. Summary of Key Sentiment Analysis Findings Neutral Dominance: The majority of responses were neutral, suggesting a lack of strong emotional responses towards not using the tool. Positive Indications: A smaller portion of positive responses indicated satisfaction with alternative methods. Lack of Negative Feedback: The absence of negative sentiments implies that the tool is not perceived negatively, but rather as unnecessary for certain users. These insights suggest that further investigation into the alternative methods used by respondents and highlighting the unique benefits of the tool could help in increasing its adoption. Additionally, educational efforts to demonstrate the tool's value might change the perception of those who currently see no reason to use it. Thematic analysis The thematic analysis resulted in comments fitting into 10 themes. Table 1 shows the themes, number of comments within the theme and three example statements from that theme. Table 1. Themes, Number of Responses, Examples Theme Number of Responses Examples Perception of Unnecessity 252 I did not need to use it. I already knew where to go. I did not feel the need to use it. Lack of Interest or Motivation 41 Did not want to. Not interested. Haven’t been interested in it. External Issues 19 I am too nervous to use it. Lack of time. I didn’t get a chance. Forgetfulness or Awareness 42 I had forgotten about spark. I'm not sure what Spark is. I am not familiar with it. Lack of Familiarity with AI Tools 37 I'm not used to it. I was not familiar. Not familiar with it. Technical or Usability 30 Not sure how to use. It was complicated. Didn't fully understand how to use it or its purpose Preference for Traditional Methods 28 I have my routine. Used textbook and other resources. I don't want to use AI. Used Other Tools 14 Used other resources. I used ChatGPT. I use a different AI tool Not Helpful 7 I tried and didn't like it. I prefer google, or bin I don't have time for the question/answer part of Spark. I am also very nervous about citing Spark.g AI copilot. Do not care for it. Other 114 Not sure. No reason, but will use it moving forward. No specific reason. I am still learning my way around the resources Perception of Unnecessity, Lack of Interest, and Lack of Awareness were the top three specific themes of why students did not utilize the AI course assistant other than the Other theme. See Figure 4 for the comparison of the amount of statements in each theme. Perception of Unnecessity was the largest theme with 252 responses. The most common reason students did not use the AI course assistant is that they did not feel it was necessary. Many students believed the resources provided by their instructors were sufficient or that they could manage their coursework without additional assistance. Typical comments include,:” I did not need to use it,” “I already knew where to go,” and “I did not feel the need to use it.” There were 41 comments that fit the category of “Lack of Interest.” A significant number of students were simply not interested or motivated to use the AI assistant. This lack of engagement could be due to various factors, including personal preference or skepticism about the tool's effectiveness. Typical comments that fit this category were, “Did not want to,” “Not interested,” and “Haven’t been interested in it.” There were 42 comments related to the theme of "Forgetfulness" and "Awareness." In other words, some students forgot about the AI assistant or were not aware of its existence. Typical comments that fit this category were, “I had forgotten about Spark,” “I'm not sure what Spark is,” “I am not familiar with it.” The thematic analysis of student comments regarding their non-use of the AI course assistant revealed several significant insights into these three categories. The most prevalent theme was a Perception of Unnecessity, with many students expressing that they did not find the AI assistant necessary for their coursework. This suggests that the resources, other than Spark, provided by instructors were perceived as sufficient or that students were confident in their ability to manage without additional help. This finding also highlights that the students did not understand that using the AI course assistants can help in learning the content more quickly and making the information more durable (Reuell, 2019). A notable number of students also cited a Lack of Interest or Motivation in using the AI assistant. This disinterest may stem from a lack of perceived value or skepticism about the tool’s effectiveness. External Issues such as time constraints and technical difficulties further hindered adoption, indicating that usability and compatibility are critical areas for improvement. This connects to and supports the findings of Al-Abdullatif, 2023 and Grassini, 2023 where they found that a students perception and familiarity with technological tools play a direct role in the students confidence to use the tool and willingness to use the tool. Forgetfulness or Awareness Issues were also identified, with some students either forgetting about the AI assistant or not being aware of its existence. This suggests that the implementation of AI tools in educational settings may require more than just an initial introduction. It highlights the need for enhanced communication strategies that go beyond a single announcement. Regular reminders through multiple channels, such as email, learning management systems, and in-class announcements, could ensure that students remain aware of the tools available to them. Forgetfulness may also stem from cognitive overload or competing priorities, especially in a busy academic environment. To mitigate this, integrating the AI assistant more seamlessly into the daily routines and course activities of students could make it a more natural part of their learning experience. For example, instructors could design assignments or in-class exercises that explicitly require the use of the AI assistant, thereby reinforcing its value and ensuring students become more familiar with its functions. Targeted communication could help address awareness issues. For instance, emphasizing the benefits of using the AI assistant in terms of academic performance and time management may encourage more students to explore and use the tool consistently. Personalized messages that demonstrate how the AI assistant can be tailored to meet individual learning needs may also be effective in driving engagement (Rogers, 2003). These strategies align with broader principles of change management in education, where continuous engagement and clear messaging are critical for the successful adoption of new technologies (Rogers, 2003). Further research could explore the specific factors that contribute to forgetfulness or lack of awareness and assess the effectiveness of various communication strategies in improving the consistent use of AI tools in educational settings. Lack of Familiarity with AI Tools was another significant barrier, as some students expressed hesitation to use the AI assistant due to limited prior experience with such technology or discomfort with learning new tools. This issue highlights the need for targeted interventions to bridge the gap between students and the effective use of AI in their educational journey. One reason for this hesitancy may be the general anxiety associated with adopting new technologies, especially among students who may already feel overwhelmed by the demands of their coursework. Jia and Tu suggest that familiarity breeds confidence, and providing students with early and frequent opportunities to engage with AI tools in a low-pressure environment could reduce their apprehension (2023). For instance, offering optional workshops, tutorials, or orientation sessions at the beginning of the term could introduce students to the AI assistant in a structured and supportive way. Ongoing support is crucial for the continued use and adoption of the AI tools. Instructors and support staff could offer regular check-ins or office hours dedicated to helping students troubleshoot any issues they encounter with the AI tools. Peer mentoring programs could also be an effective strategy, where more tech-savvy students assist their peers in learning and using the AI assistant. This approach not only fosters a collaborative learning environment but also encourages the sharing of best practices among students. Integrating AI tools into course activities and assignments gradually, rather than introducing them all at once, could help students build their proficiency over time. For example, starting with simple tasks that require the use of the AI assistant and gradually increasing complexity can make the learning curve less steep. A small group of students preferred Traditional Methods of studying, relying on established routines or other recommended programs. This preference indicates that the AI assistant needs to clearly demonstrate its unique benefits to convince students of its added value. Summary of Results The sentiment and thematic analyses together reveal that the primary barriers to using the AI course assistant are practical in nature rather than rooted in strong negative sentiments. The predominance of neutral responses underscores a perception of unnecessity, a lack of interest, and forgetfulness or unawareness as key factors. The sentiment and thematic analyses of student responses to the question regarding their non-use of the AI course assistant, Spark, reveal several key insights. Sentiment Analysis Findings Neutral Sentiment Dominance: The majority of responses (72.3%) exhibited neutral sentiment, indicating a lack of strong emotional responses towards not using the tool. Students commonly expressed practical or situational reasons, such as "I had no reason to use it." Positive Sentiment: A smaller portion of responses (27.7%) were positive, reflecting confidence in alternative methods or satisfaction with other tools. Comments such as "I was able to do research on my own" illustrate this sentiment. Absence of Negative Sentiment: No negative sentiment was detected, suggesting that the tool is not viewed unfavorably, but rather as unnecessary for some students. Thematic Analysis Findings Perception of Unnecessity: The most common theme, with 252 responses, was the belief that the AI assistant was not needed. Students felt that the resources provided by their instructors were sufficient or that they could manage their coursework without additional help. Lack of Interest or Motivation: A significant number of students (41 responses) expressed disinterest in using the AI assistant, often due to personal preference or skepticism about its effectiveness. Forgetfulness or Awareness Issues: Some students (42 responses) forgot about the AI assistant or were unaware of its existence, highlighting the need for improved communication and regular reminders. Lack of Familiarity with AI Tools: Another barrier identified was students' discomfort with new technology, underscoring the importance of training and support to help students become more comfortable with AI tools. External Issues and Traditional Methods: Other factors such as external pressures, technical difficulties, and a preference for traditional study methods also contributed to the non-use of the AI assistant. The results indicate that most students did not perceive a strong need for the AI assistant, likely due to the adequacy of existing resources or confidence in their own abilities. The lack of negative sentiment suggests that the tool itself is not seen as problematic, but rather that its perceived necessity and value may need to be communicated more effectively. Addressing issues of awareness, familiarity, and support will be crucial in increasing the adoption of AI tools in educational settings. These findings suggest that with the right interventions, the AI assistant has the potential to be a valuable resource for more students, particularly if efforts are made to address the barriers identified in this study. Discussion The findings from the sentiment and thematic analyses provide significant insights into students' perceptions and use of AI course assistants like Spark. The predominant theme of perceived unnecessity suggests that students generally do not view these tools as essential for their academic success. This perception aligns with the sentiment analysis, where a large proportion of neutral responses indicated a practical rather than emotional rationale for non-use. In both of these cases, it is apparent that students are not aware of the value of using tools such as Spark to engage in active learning. Students learn more with active learning, yet they enjoy the process less and believe they learn more with traditional lectures (Reuell, 2019). In light of this data, we are more able to focus our efforts towards specific areas to help students engage with Spark and achieve greater learning through engaging in active learning with Spark. This suggests a need for better communication and reminders about the availability and benefits of using the tool. Practicality Over Necessity The overwhelming sentiment of neutrality indicates that many students do not perceive the AI course assistant as a necessary tool in their learning process. This suggests a possible gap between the potential benefits of AI course assistants and students' understanding or experience of these benefits. As Reuell (2019) indicated, active learning environments enhance learning retention and outcomes, yet students may not see how AI tools like Spark contribute to this active learning process. The findings indicate a need for better communication of how these tools can enhance learning efficiency and support academic success. Lack of Engagement and Awareness The theme of Lack of Interest or Motivation suggests that some students are not compelled to use AI assistants due to a lack of perceived value or skepticism about their effectiveness. Moreover, the themes of "Forgetfulness or Awareness" and "Lack of Familiarity with AI Tools" highlight the importance of increasing awareness and understanding of AI capabilities. Many students indicated forgetfulness or unfamiliarity with Spark, suggesting that enhanced communication and training could address these barriers. Technical and Usability Concerns The analysis also pointed out technical or usability issues as a barrier to adoption. While not the most prominent reason for non-use, the existence of these concerns emphasizes the need for continuous improvement of AI tools to ensure the tools are user-friendly and compatible with various student needs and technological proficiencies. This aligns with the need for training and support systems that can help bridge the gap between students and technology, making the transition smoother and more accessible. Traditional Methods and Resistance to Change The preference for traditional methods, as noted in the thematic analysis, underscores a resistance to change that is common in educational settings. Students' reliance on established routines and resources suggests that AI tools need to demonstrate clear and unique benefits to be considered valuable additions rather than mere alternatives. It is crucial to highlight these features of AI course assistants that distinguish them from traditional methods, potentially increasing their attractiveness and adoption among students. Recommendations Based on these findings, several strategies are recommended to increase the adoption and effectiveness of AI course assistants: Enhanced Communication and Training: Institutions should focus on raising awareness for both students and faculty about AI tools through informational sessions, demonstrations, and success stories to illustrate their benefits. Providing training sessions to familiarize students and faculty with the technology can alleviate hesitations related to unfamiliarity.l This would come in the form of a series of short videos (less than 90 seconds) that demonstrate capabilities and create awareness. Videos would be delivered via emai, course pop-ups, and within the universities app. Highlighting Unique Benefits: Clearly communicate the specific advantages of AI course assistants over traditional methods, such as personalized learning paths and real-time feedback, to enhance learning experiences. Engaging Students: Increase motivation by integrating AI tools into regular coursework and showcasing their effectiveness through practical examples and testimonials. Provide Training and Support : Early and ongoing support, including workshops and peer mentoring, can help students become more comfortable and proficient with AI tools. Further Research Future research should explore the long-term impact of AI assistants on student learning outcomes and the specific features that most effectively drive engagement and success. By doing so, institutions can fully realize the benefits of AI integration in education, ultimately supporting students in achieving their academic goals. Another important avenue for research is examining how professors' utilization and buy-in of AI course assistants in educational settings influence student use and engagement with these tools. Conclusion Addressing these barriers is not merely about increasing tool adoption; it is about ensuring equitable access to transformative learning technologies that can bridge educational gaps, enhance lifelong learning, and prepare students for an increasingly AI-integrated society. AI course assistants represent a promising avenue for enhancing educational experiences. By understanding and addressing the perceptions and barriers to their adoption, educational institutions can harness the full potential of these tools, ultimately creating a more dynamic and personalized learning environment. This will require a concerted effort to increase awareness among students and faculty, provide training, and continuously improve the usability of AI tools to ensure their successful integration into academic settings. This study explored the factors contributing to some students' non-utilization of the AI course assistant Spark by students at Los Angeles Pacific University, as well as strategies to enhance Spark’s utilizaton and engagement by students.. Through the analysis of student feedback, several key insights were revealed that address our research questions. RQ1: "What are the primary factors (e.g., technological, psychological, educational) that contribute to some students not utilizing the AI course assistant Spark available in their online classroom?" The findings indicate that the predominant barrier is a perception of unnecessity, where students feel that existing resources are sufficient and do not perceive additional value in using the AI assistant. Lack of interest or motivation and forgetfulness or unawareness also emerged as significant factors, suggesting that many students are not fully aware of the assistant's capabilities or benefits. Additionally, a preference for traditional study methods and technical usability issues were identified as barriers, highlighting the psychological and educational aspects influencing student decisions. RQ2: "What strategies can be implemented to increase the utilization of the AI course assistant Spark among students, thereby enhancing their learning experience and academic performance?" The study suggests several actionable strategies. Enhancing communication and training about the AI assistant's features and benefits can address issues of awareness and familiarity. Highlighting the unique benefits of the AI assistant, such as personalized learning and real-time feedback, can motivate students by demonstrating clear advantages over traditional methods. Engaging students through integration of the AI assistant into regular coursework and showcasing its effectiveness with practical examples can further encourage adoption. While the AI course assistant Spark offers significant potential to enhance learning experiences, its adoption is hindered by perceptions of unnecessity and a lack of awareness. By addressing these barriers through targeted strategies, educational institutions can better leverage AI tools to create more dynamic and personalized learning environments. Declarations The research was ethically approved by the Institutional Review Board (IRB) of Los Angeles Pacific University. Prior to participation, all participants were duly informed of their rights and responsibilities and provided explicit written consent. The study was conducted in agreement with the guidelines governing research involving human participants, as outlined by the IRB . Author Contribution The manuscript was a collaborative piece between G.H. and C.S. Both authors contributed to each section of the paper with one author being primarily responsible. G.H. and C.S. analyzed the data separately and as a team. Data gathering was a primary responsibility of G.H. G.H. was the primary writer of the data analysis and discussion sections. C.S. wrote the literature review. Both G.H. and C.S. wrote the introduction and conclusion sections. Both authors reviewed the manuscript. Data Availability Data is provided at this link https://docs.google.com/spreadsheets/d/1I9rkVPzO1L-C91rpS-Q2WLGJh3U0SjaJcMR82mSu-2Q/edit?usp=sharing References Al-Abdullatif, A. M. (2023). Modeling students’ perceptions of chatbots in learning: Integrating technology acceptance with the value-based adoption model. Education Sciences , 13 (11), 1151. https://doi.org/10.3390/educsci13111151 Chen, X., Zou, D., Cheng, G., & Xie, H. (2022). AI in education: A review of current research and applications. Journal of Educational Technology & Society, 25 (1), 31-44. Deng, X., & Yu, Z. (2023). A meta-analysis and systematic review of the effect of chatbot technology use in sustainable education. Sustainability , 15 (4), 2940. https://doi.org/10.3390/su15042940 Essel, H. B., Vlachopoulos, D., Tachie-Menson, A., Jhonson, E. E., Baah, P. K. (2022). The impact of a virtual teaching assistant (chatbot) on students' learning in Ghanaian higher education. International Journal of Educational Technology in Higher Education , 19 (1), 57. https://doi.org/10.1186/s41239-022-00362-6 Grassini, S. (2023). Shaping the future of education: Exploring the potential and consequences of AI and ChatGPT in educational settings. Education Sciences , 13 (7), 692. https://doi.org/10.3390/educsci13070692 Hanshaw, G., & Miller, K. (2024). Evaluating the impact of real-time AI feedback on student writing: Randomized control trial. Manuscript in preparation. Los Angeles Pacific University. Ilieva, G., Yankova, T., Klisarova-Belcheva, S., Dimitrov, A., Bratkov, M., & Angelov, D. (2023). Effects of generative chatbots in higher education. Information , 14 (9), 492. https://doi.org/10.3390/info14090492 Labadze, L., Grigolia, M., & Machaidze, L. (2023). Role of AI chatbots in education: Systematic literature review. International Journal of Educational Technology in Higher Education , 20 (1), 1-17. https://doi.org/10.1186/s41239-023-00426-1 Luckin, R., Holmes, W., Griffiths, M., & Forcier, L. B. (2016). Artificial intelligence in education: Promises and implications for teaching and learning. Learning, Media and Technology, 41(1), 1-19. https://doi.org/10.1080/17439884.2016.1139939 Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879-903. https://doi.org/10.1037/0021-9010.88.5.879 Reuell, P. (2019, September 4). Study shows that students learn more when taking part in classrooms that employ active-learning strategies . Harvard Gazette. https://news.harvard.edu/gazette/story/2019/09/study-shows-that-students-learn-more-when-taking-part-in-classrooms-that-employ-active-learning-strategies/ Rogers, E. M. (2003). Diffusion of innovations (5th ed.). Free Press. Williams, R. T. (2024). The ethical implications of using generative chatbots in higher education. Frontiers in Education , 8 , 1331607. https://doi.org/10.3389/feduc.2023.1331607 Wu, R., & Yu, Z. (2023). Do AI chatbots improve students’ learning outcomes? Evidence from a meta‐analysis. British Journal of Educational Technology , 55 (1), 10-33. https://doi.org/10.1111/bjet.13334 Additional Declarations No competing interests reported. Supplementary Files AppendixAtoE.docx Cite Share Download PDF Status: Published Journal Publication published 25 Jul, 2025 Read the published version in Discover Artificial Intelligence → Version 1 posted Editorial decision: Revision requested 12 Mar, 2025 Reviews received at journal 26 Feb, 2025 Reviews received at journal 24 Feb, 2025 Reviews received at journal 22 Feb, 2025 Reviewers agreed at journal 21 Feb, 2025 Reviews received at journal 21 Feb, 2025 Reviewers agreed at journal 18 Feb, 2025 Reviewers agreed at journal 18 Feb, 2025 Reviewers agreed at journal 18 Feb, 2025 Reviewers agreed at journal 18 Feb, 2025 Reviewers invited by journal 16 Feb, 2025 Editor assigned by journal 11 Feb, 2025 Submission checks completed at journal 06 Feb, 2025 First submitted to journal 20 Jan, 2025 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-5867866","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":412121271,"identity":"7b487c22-e1ad-4428-bd30-b714224da7f4","order_by":0,"name":"George Hanshaw","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAyklEQVRIiWNgGAWjYDACCSBmbGBg4GeH8EFsIrVINpOsxeAwsVrko5ufPfy5wy5x82HmZ495GGxkNxwgoMXwzjFzY94zyYnbDrOZG/MwpBkT1jIjwUyasY05d9thBjNpHobDiURoSf8m+bOtPndzM/s3oJb/hLXIS+SYSfC2Hc7dwMwDsuUAYS0GEjll0rxtx+tnHOYpN5xjkGw8k6AtM9K3AR1Wbczf3r7twZsKO9k+grYgKWADcgkoB9vSgKJlFIyCUTAKRgEWAACPVkGUr3zBdwAAAABJRU5ErkJggg==","orcid":"","institution":"Los Angeles Pacific University","correspondingAuthor":true,"prefix":"","firstName":"George","middleName":"","lastName":"Hanshaw","suffix":""},{"id":412121272,"identity":"b5b30a5f-aaef-412d-93f3-0e2247ea530e","order_by":1,"name":"Christopher Sullivan","email":"","orcid":"","institution":"Los Angeles Pacific University","correspondingAuthor":false,"prefix":"","firstName":"Christopher","middleName":"","lastName":"Sullivan","suffix":""}],"badges":[],"createdAt":"2025-01-20 17:23:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5867866/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5867866/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s44163-025-00312-x","type":"published","date":"2025-07-25T15:58:09+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":75896267,"identity":"3418117c-3ac1-43cd-bb06-a77e70f75f35","added_by":"auto","created_at":"2025-02-10 10:32:43","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":66292,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eDistribution of Sentiment Scores\u003c/em\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5867866/v1/2e52a4f3ef64324347173d0a.png"},{"id":75895779,"identity":"366c62fd-7e9d-4b9b-83cd-b9ae6c7d95ef","added_by":"auto","created_at":"2025-02-10 10:24:43","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":27688,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eProportion of Responses\u003c/em\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5867866/v1/bc8e7368ea258c1ad1ad6d55.png"},{"id":75896268,"identity":"802c9035-c678-4102-a004-6e253dc286e7","added_by":"auto","created_at":"2025-02-10 10:32:43","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":447143,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eWord Cloud of Survey Responses\u003c/em\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5867866/v1/22618a5a0d4f3cb5523cdda1.png"},{"id":75895782,"identity":"a889382d-848d-4b09-aa45-af0686823123","added_by":"auto","created_at":"2025-02-10 10:24:43","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":98862,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eStatements in Each Theme\u003c/em\u003e\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-5867866/v1/d258ad2072952043f98a1311.png"},{"id":88506336,"identity":"76903760-b6b2-4c89-938a-15a5af531dd0","added_by":"auto","created_at":"2025-08-07 07:32:48","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1648392,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5867866/v1/30e5e7b0-33d4-4f2f-bb8a-b07beadc766c.pdf"},{"id":75895780,"identity":"1897f0fd-5a15-436d-ad60-cf68fa938039","added_by":"auto","created_at":"2025-02-10 10:24:43","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1724920,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixAtoE.docx","url":"https://assets-eu.researchsquare.com/files/rs-5867866/v1/257233d301184c152ae6ded4.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Exploring Barriers to AI Course Assistant Adoption: A Mixed-Methods Study on Student Non-Utilization","fulltext":[{"header":"Why Students Did Not Use AI Course Assistants","content":"\u003cp\u003eIn recent years, the integration of artificial intelligence (AI) in educational settings has shown tremendous potential in enhancing student learning experiences. AI-powered tools, such as course assistants, have been designed to provide personalized learning, immediate feedback, and adaptable instructional support that aligns with the needs of modern learners. Research has demonstrated that AI can significantly improve student engagement, motivation, and overall academic performance by offering tailored educational experiences (Chen, Zou, Cheng, \u0026amp; Xie, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Deng \u0026amp; Yu, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Hanshaw et al. (2024) found that students who use AI course assistants designed to employ the Socratic method, encouraging inquiry without completing assignments for them, achieve significantly higher grade point averages compared to students who do not use these tools. Wu and Yu (2024) found that students who utilize AI chatbots within their courses achieve better grade outcomes as well. Understanding and addressing the barriers to adoption, these tools risk becoming underutilized resources, exacerbating inequities in educational outcomes, and leaving many students under-prepared for a rapidly evolving technological landscape.\u003c/p\u003e \u003cp\u003eDespite these benefits, the actual utilization of such AI tools by students remains inconsistent, with a significant portion opting not to engage with these resources. This paradox between the potential benefits and the actual usage of AI in education underscores a significant gap in our understanding of how students utilize AI. Previous studies have indicated that students' perceptions of AI, their familiarity with such technologies, and their confidence in using them play a significant rolein their adoption of these tools (Al-Abdullatif, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Grassini, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThis study focuses on the AI course assistant Spark used at Los Angeles Pacific University (LAPU), which was introduced to support students in their courses and other university-level resources such as the writing hub. Preliminary findings from a pilot study indicate that students who utilized Spark in their courses demonstrated higher academic performance, motivation, and efficacy compared to non-users (Hanshaw et. al, 2024). However, a substantial number of students reported not using the AI assistant, raising critical questions about the factors contributing to this non-use.\u003c/p\u003e \u003cp\u003eUnderstanding these factors is vital for maximizing the potential of AI in education generally, and for the adoption of AI course assistants specifically. If AI tools such as course assitants are underutilized, their ability to enhance educational outcomes is severely compromised. Therefore, this study aims to explore the technological, psychological, and educational reasons behind students' non-utilization of Spark and to develop strategies that could increase its adoption, thereby enhancing the overall learning experience.\u003c/p\u003e \u003cp\u003eBy employing a mixed-methods exploratory approach, this research seeks to identify both the barriers to and the opportunities for increasing AI course assistant usage among students. The insights gained will not only contribute to the literature on AI in education, but will also provide practical recommendations for educators and institutions aiming to foster greater engagement with AI technologies (Ilieva et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Labadze, Grigolia, \u0026amp; Machaidze, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e"},{"header":"Literature Review","content":"\u003cp\u003eAI is a driving force behind many technologies used daily throughout society, such as household gadgets, fitness monitors, self-driving cars, and social media platforms. In educational settings, AI is revolutionizing the student learning experience (Labadze et al., 2023; Ilieva et al., 2023; Luckin et al. 2016). Despite these benefits, some students in our study did \u0026nbsp;not use the AI course assistant, Spark, in their online classrooms. This literature review explores the reasons behind this underutilization, focusing on issues of awareness and interest.\u003c/p\u003e\n\u003cp\u003eFor students in their college years, who have been raised in an era marked by swift technological progress, AI is not an unfamiliar concept. These students have been brought up with AI-enabled tools and platforms, ranging from tailored learning applications to social media algorithms. (Hyesun et al., 2023) Still, adoption of AI in higher education is just beginning, \u0026ldquo;Data\u0026hellip;shows a dramatic rise in AI usage since 2020 with a 150% increase from the prior 2 years 2020\u0026ndash;2019.\u0026rdquo; (Crompton \u0026amp; Burke, 2023), yet students are not always as eager to engage with and adopt AI technology in education as they do with the other \u0026ldquo;Smart\u0026rdquo; interfaces they interact with regularly.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe incorporation of AI in the realm of education has its roots in the influential work of Vygotsky (1978), particularly his theory of the Zone of Proximal Development (ZPD). This theory suggests that effective learning takes place in a zone where learners are sufficiently challenged within their capabilities yet still require guidance and support. AI tailors educational activities to the learners\u0026rsquo; \u0026nbsp;readiness level, ensuring they are neither understimulated nor overwhelmed, thus optimizing the learning experience. AI also possesses the ability to track the progression of the learning process, adapting in real-time as the learner interacts with the system. This continuous feedback loop enables the system to modify the learning activities to align with the learner\u0026rsquo;s evolving progress and needs.\u003c/p\u003e\n\u003cp\u003eIn the context of online education, adaptive systems employ AI to cultivate self-regulated learning among students. This approach encompasses a spectrum of learning abilities, including goal setting, self-monitoring, self-instruction, and self-reinforcement (Ng et al., 2023). Moreover, the incorporation of AI in differentiated learning has transformed personalized education. AI now equips educators with insights into students\u0026rsquo; learning strategies, backgrounds, progress, and academic interests. Significantly, AI can aid in bridging the educational divide caused by inequality issues, thus fostering accessibility for all learners, including those with special educational needs (Ng et al., 2023).\u003c/p\u003e\n\u003cp\u003eThe evolution of technology in education has undergone a transformative journey. The earliest \u0026ldquo;teaching machines\u0026rdquo; were designed to facilitate learning in a manner reminiscent of a private tutor. In the 1950s, a significant advancement was made with the development of Self Adaptive Keyboard Instructors (SAKI). These systems were designed to adapt to student performance but could not be adjusted to suit the needs of individuals. A significant milestone in educational technology was reached in 1970 when AI was first applied to these educational systems, leading to the creation of Intelligent Tutoring Systems (ITS) (Tonbuloglu, 2023). These systems represented a significant leap forward in the field of educational technology, laying the foundation for the sophisticated, AI-driven learning platforms we see today. Despite their potential, these systems did not gain widespread recognition due to cost and accessibility issues.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eApplications\u0026nbsp;\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eAI has brought about a significant shift in the educational sector. The use of AI in education can enable the development of more personalized curricula on a global scale, making it relatively economical not only in terms of monetary investment but also in the time and attention required from educators (Tonbuloglu, 2023). AI plays a critical role in personalized learning experiences based on each individual\u0026rsquo;s needs and learning styles by identifying patterns in students\u0026rsquo; learning behaviors, preferences, and achievements. Additionally, AI-enabled virtual assistants and chatbots in an online classroom can provide instant support and guidance to students, answering questions and analyzing student performance. These abilities of AI course assistants to accomplish these tasks has been shown to have a positive impact on student learning (Essel et al., 2022).\u003c/p\u003e\n\u003cp\u003eIn addition to personalized learning experiences through AI algorithms, instructors can analyze learners\u0026rsquo; data to predict student performance and provide additional support. By leveraging this data, AI has reliably become a tool to predict performance and identify those at risk of failure, allowing instructors to intervene and provide further assistance. This allows for instructors to identify areas for improvement and tailor their teaching strategies accordingly, adjusting learning content and leveraging the use of teaching assistants and chatbots to provide immediate feedback and personalized testing.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe advent of big data and AI has significantly reshaped management and higher education practices and is currently applicable within the following scenarios (Jia \u0026amp; Tu, 2023):\u0026nbsp;\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cstrong\u003eAssessment of Students and Schools\u003c/strong\u003e: These methods employ academic analytics to deliver a more customized educational experience.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eGrading and Evaluation\u003c/strong\u003e: Technologies such as image recognition, computer vision, and prediction systems have streamlined the grading and evaluation of papers and exams.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003ePersonalized Intelligent Teaching\u003c/strong\u003e: Intelligent teaching systems and learning analytics are employed to adapt teaching methods to individual student needs.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eSmart Facilities\u003c/strong\u003e: The incorporation of facial recognition, speech recognition, virtual labs, augmented reality (A/R), virtual reality (V/R), and hearing and sensing technologies.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eOnline and Mobile Remote Education\u003c/strong\u003e: The utilization of edge computing, virtual personalized assistants, and real-time analysis, thusmaking learning more accessible and flexible.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2\u003e\u003cstrong\u003eStudents \u0026amp; Teachers\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe use of AI capabilities has had a noticeable impact on students, not only in providing personalized learning experiences, but also in various aspects such as interactivity, real-time feedback, and customization and improvement of resource management in teaching and learning environments (Jia \u0026amp; Tu, 2023). These customized learning environments may better stimulate students\u0026rsquo; motivation to learn, as they perceive the alignment between learning content and their individual needs.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdditionally, the interactivity and real-time feedback facilitated by AI applications can offer personalized pacing and error correction, allowing students to experience their learning progress more immediately. Such immediate feedback mechanisms can enhance students\u0026rsquo; self-efficacy and consequently boost their motivation to learn (Crompton \u0026amp; Burke, 2023). It also provides a unique ability to scale to model the student\u0026apos;s learning process, determine the necessary information through performance analysis, make decisions, and provide interaction by responding to student questions and directing questions to them. (Tonbuloglu, 2023)\u003c/p\u003e\n\u003cp\u003eSome of the features that make the use of AI in educational processes different from other technologies are its ability to model the student\u0026apos;s learning process, to determine the information needed through performance analysis and to make decisions, and to provide interaction by responding to student questions and directing questions to them (Tonbuloglu, 2023). AI-driven systems can develop custom learning profiles for each student and customize their learning journeys and materials based on their needs, ability, preferred mode of learning, and experience (Ng et al., 2023).\u003c/p\u003e\n\u003cp\u003eThe affordances of AI can also ensure that the support is timely without waiting for a person to be available. Also, AI course assistants can consider aspects of students\u0026rsquo; academic ability, preferences, and best strategies for support (Crompton \u0026amp; Burke, 2023).\u003c/p\u003e\n\u003cp\u003eAI has become an important tool for educators. AI-powered tools have evolved to be more educator-centric, helping teachers identify effective teaching methods based on students\u0026apos; learning data. These tools also automate administrative tasks, create assessments, and handle grading and feedback (Ng et al., 2023). This saves teachers time and increases efficiency. AI can assist teachers by generating recurring questions, providing students with personalized assistance, and facilitating communication with peers. Through data analysis, teachers can adjust their teaching approaches and customize learning resources to meet students\u0026apos; needs (Ng et al., 2023). Teachers can adapt their methods to changing learning scenarios and objectives, whether in traditional classrooms or online platforms. AI can also improve the teaching and learning process by transforming instructional design, evaluation, and learning environments (Tonbuloglu, 2023).\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eTrust and Usefulness\u0026nbsp;\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eA user\u0026apos;s perception of trustfulness and usefulness play an important role in how people / students interact with AI. Trust is an essential human mechanism that helps us navigate vulnerability, uncertainty, complexity, and ambiguity, all of which make up risk. It is \u0026nbsp;a psychological state that enables us to be vulnerable based on the optimistic expectations of an outcome or someone\u0026rsquo;s actions. In contrast to conventional technologies that operate based on user commands and predetermined rules, AI operates with a level of independence. The unpredictable nature of AI, often referred to as a \u0026quot;black-box,\u0026quot; underscores the critical importance of trust as users confront the complexity and potential hazards of AI\u0026rsquo;s decision processes (Hyesun et al., 2023). Within the realm of technology, trust is built upon three foundational dimensions: benevolence/helpfulness, integrity/reliability, and capability/effectiveness.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe Technology Acceptance Model (TAM) initially posited that the two primary drivers of a user\u0026apos;s intention to continue using technology are its perceived usefulness and ease of use. According to subsequent findings, trust enhances the perceived usefulness, which in turn boosts the likelihood of continued use. Trust also cultivates positive attitudes, further influencing the intention to use the technology. Moreover, trust has been shown to increase the perceived usefulness, which leads to more positive attitudes and a higher intention to use the technology. (Hyesun et al., 2023)\u003c/p\u003e\n\u003cp\u003eAs technology continues to advance, AI has emerged as a key player, introducing new dimensions of trust that go beyond traditional technologies. For AI, trust can be looked at from two perspectives: trust in AI\u0026apos;s human-like qualities (essentially, the personality of the technology), and trust in the operational aspects of AI (its capability, dependability, and safety). However, perceived usefulness extends beyond trust alone. Users assess usefulness in terms of how effectively AI can meet their needs, enhance productivity, and deliver tangible benefits that justify its integration into their daily routines. When users perceive AI as genuinely useful, they are more likely to overlook potential risks and invest the necessary time and effort to learn and adapt to the technology.\u003c/p\u003e\n\u003cp\u003eThese facets, both in AI\u0026apos;s human-like qualities and its operational aspects, greatly affect individuals\u0026rsquo; perceptions of smart technology\u0026apos;s usefulness and appeal, thereby influencing their willingness to use it. Cognitive trust is influenced by factors such as transparency, dependability, and the nature of the task, whereas emotional trust is shaped by assigning human characteristics to AI. Of all the factors, perceived ease of use had the most substantial overall impact on usage, followed by perceived usefulness and trust. (Hyesun et al., 2023)\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eOther barriers\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eWhile the implementation of AI in administrative and educational settings aims to improve efficiency and automate processes, its direct impact on students\u0026apos; critical thinking abilities is uncertain. Critical thinking goes beyond quick information processing and requires skills for independent thought, thorough understanding and sound judgment. Although AI is proficient at analyzing large amounts of data, it primarily involves inputting information. On the other hand, critical thinking involves deep evaluation, reasoning, and critiquing of information, which depend not just on AI but also on learning, reflection, and practical experience (Jia \u0026amp; Tu, 2023).\u003c/p\u003e\n\u003cp\u003eThere are also remaining issues about how this change in social interaction, particularly with the increasing presence of AI, plays out in developing additional skill sets related to relationships and thinking. While AI can assist in various educational tasks, its role in fostering social skills and self-efficacy is more complex. Research has shown that having beliefs in one\u0026apos;s own ability to achieve goals, known as general self-efficacy, plays a crucial role in mental well-being, physical health, and the ability to change behavior (Jia \u0026amp; Tu, 2023). Social interaction, traditionally a human-centered activity, is fundamental in the development of self-efficacy. Through engaging in various social environments and interactions, individuals can observe, experience, and practice behaviors, which in turn reinforce their beliefs in their capabilities.\u003c/p\u003e\n\u003cp\u003eAs AI increasingly mediates social interactions, the dynamics of how self-efficacy is developed may shift. While AI can provide feedback and simulate certain social scenarios, it may lack the depth of human interaction needed to fully foster self-efficacy. Bandura\u0026apos;s (1995) research has highlighted the influence of personal beliefs in one\u0026apos;s abilities within social and cultural contexts, shaping individuals\u0026apos; life paths. The feedback, modeling, and encouragement received from others in social settings significantly contribute to strengthening one\u0026apos;s self-efficacy, and AI\u0026apos;s ability to replicate this may be limited.\u003c/p\u003e\n\u003cp\u003eOther studies have emphasized the significance of life and career skills such as problem-solving, emotional intelligence, judgment, service orientation, negotiation, cognitive flexibility, as well as communication and teamwork skills in the fourth industrial revolution. AI can support the development of some of these skills, particularly in problem-solving and cognitive flexibility, but the human elements of communication, emotional intelligence, and teamwork require direct human interaction and intervention. Developing critical thinking through academic reading, another area AI can support, is essential for meeting the higher-order thinking requirements of 21st-century students (Ng et al., 2023). Therefore, while AI has its place in education, it is clear that human skills and intervention remain vital.\u003c/p\u003e\n\u003cp\u003eSeveral potential risks and conflicts, such as privacy concerns, changes in how power is structured, and excessive control over programming and data have been identified bystudents and teachers due to the potential of creating misunderstandings or misleading information. There are concerns that AI could provide unreliable recommendations, which may negatively impact students\u0026apos; performance, especially if teachers depend solely on AI-driven technologies to predict and assess students\u0026apos; learning outcomes (Ng et al., 2023). It\u0026apos;s important to recognize that AI-driven platforms can sometimes misinterpret users and provide inaccurate suggestions. Therefore, student learning outcomes and social interactions should not rely solely on AI interpretation.\u003c/p\u003e\n\u003cp\u003eLastly, the design of AI driven tools may not be sufficiently human-centered (or even student-centered), which may cause discomfort for students. For example, features such ash as eye tracking or facial expression analysis may feel like surveillance to students. AI-based misunderstandings, misleading information, limitations, and hidden ethical issues have been noted by researchers and experts in the field (Ng et al., 2023).Therefore, AI-competency for teachers is essential to enhance students\u0026apos; AI-driven online learning, and teachers need to enhancetheir skills and knowledge through continuous professional development. \u0026nbsp;Most notably, \u0026ldquo;The researchers found a great lack in pedagogical and ethical implications of implementing AI in HE and that there was a need for more educational perspectives on AI developments from educators conducting this work\u0026rdquo; (Crompton \u0026amp; Burke, 2023).\u0026nbsp;\u003c/p\u003e\u003ch2\u003e\u003cstrong\u003eConclusion\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eAI has the potential to revolutionize educational experiences by providing personalized learning, immediate feedback, and adaptable instructional support; however, its actual utilization among students remains inconsistent. This literature review has highlighted various factors influencing the adoption of AI tools in educational settings, including technological familiarity, trust, and perceived usefulness on the part of both teachers and students. Understanding these factors is crucial for developing strategies to enhance student engagement with AI technologies, ultimately maximizing their educational benefits. What is lacking in the current body of research is an exploration of the reasons students may not engage with AI course assistants placed within their classroom. Therefore, by focusing on the AI course assistant Spark at Los Angeles Pacific University, this study aims to delve deeper into the specific reasons behind its underutilization. Understanding these factors is crucial for developing strategies to enhance student engagement with AI technologies, ultimately maximizing their educational benefits. The insights gained from this research will contribute to the broader literature on AI in education and offer practical recommendations for educators and institutions seeking to foster greater adoption of AI tools.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eResearch Question\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eRQ1: What are the primary factors (e.g., technological, psychological, educational) that contribute to some students not utilizing the AI course assistant Spark available in their online classroom?\u003c/p\u003e\n\u003cp\u003eRQ2: What strategies can be implemented to increase the utilization of the AI course assistant Spark among students, thereby enhancing their learning experience and academic performance?\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003ePurpose\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eAI, as outlined by Kim et al. (2020), is increasingly being designed to teach, interact, and adapt to human teaching and learning methods, offering significant potential to enhance educational experiences. LAPU\u0026apos;s pilot study on the use of the AI assistant Spark demonstrated that student-usersnot only outperformed non-users academically, but also showed higher levels of motivation and efficacy (Hanshaw et al., 2024). Given these clear benefits and the relatively straightforward adoption process, our research seeks to explore why some students choose not to use the AI assistant. Understanding these reasons will enable LAPU designers and faculty members to refine the design and promotion of AI course assistants, thus showcasing the value and benefits in delivering personalized and contextualized learning experiences.\u003c/p\u003e\n\u003cp\u003eAfter identifying the barriers and reasons behind why some students did not use the AI course assistant, we will identify ways to better educate and connect with students to increase the utilization rate, enabling them to reap the benefits of using AI course assistants. By addressing and mitigating these barriers, we hope to foster greater acceptance and integration of AI course assistants and AI technology in general within educational settings.\u003c/p\u003e"},{"header":"Method","content":"\u003ch2\u003eStudy Design\u003c/h2\u003e\n\u003cp\u003eThis study employs a mixed-methods exploratory approach to understand the factors influencing students\u0026apos; decisions not to use the AI course assistant Spark at LAPU. By integrating both qualitative and quantitative methods, this framework allows for a comprehensive examination of student perceptions and behaviors, providing nuanced insights into the barriers and potential strategies for increasing AI course assistant adoption.\u003c/p\u003e\n\u003ch2\u003eRationale for Mixed-Methods Approach\u003c/h2\u003e\n\u003cp\u003eThe sentiment analysis quantifies the emotional tone of student responses, providing an overview of the general attitudes towards the AI tool. This component helps to identify the prevalence of different sentiments (positive, neutral, or negative) within the student population.\u003c/p\u003e\n\u003cp\u003eThematic analysis rigorously investigates the qualitative data, pinpointing distinct themes and sub-themes that elucidate the factors contributing to disengagement. This method explores the intrinsic motivations, impediments, and situational elements that influence the students\u0026rsquo; choices.\u003c/p\u003e\n\u003ch2\u003eJustification for Exploratory Design\u003c/h2\u003e\n\u003cp\u003eThe study aims to explore and identify factors affecting the adoption of AI course assistants in an educational setting without prior hypotheses. This exploratory nature is essential for understanding the complex and multifaceted reasons students may not engage with available AI technologies. The integration of quantitative and qualitative data provides a more holistic understanding of the issue, ensuring that both statistical trends and personal narratives are considered in the analysis.\u003c/p\u003e\n\u003ch2\u003eObjectives of the Framework\u003c/h2\u003e\n\u003cp\u003eThe objective of this framework is to identify perceived or real barriers towards the use of the AI course assistants as well as to develop strategies to help students overcome these barriers and enhance the utilization of the AI course assistant. Identifying barriers requires us to uncover the technological, psychological, and educational factors that contribute to the non-use of the AI course assistant. After identifying barriers, we then formulate strategies to enhance the utilization of AI tools, thereby improving student learning experiences and outcomes.\u003c/p\u003e\n\u003ch2\u003eParticipants\u003c/h2\u003e\n\u003cp\u003eAll participants were active LAPU undergraduate and graduate students. LAPU is a fully online, accredited university that caters primarily to adult learners. With a focus on flexibility, LAPU offers a variety of programs designed to meet the needs of working professionals, parents, and individuals balancing multiple responsibilities. The student body is diverse, with learners from various backgrounds and experiences, many of whom are returning to education after a significant time away from formal learning.\u003c/p\u003e\n\u003cp\u003eA total of 883 End-Of-Course (EOC) surveys were completed and submitted by students during the Summer 1 term in 2024. Of these, approximately 68% of students self-reported not using Spark, the AI course assistant. This study specifically analyzed the responses from 602 EOC surveys where students indicated they did not use Spark.\u003c/p\u003e\n\u003cp\u003eThe EOC surveys are administered to students in every course, and participation is both optional and anonymous. It is important to note that the 602 responses do not necessarily represent 602 unique students, as some students may have completed more than one course and thus submitted multiple surveys. Each student can submit the survey only once per course.\u003c/p\u003e\n\u003cp\u003eThe overall survey response rate was 42.1%. The respondents included students in both undergraduate and graduate-level courses.. While the demographic composition of the respondents is believed to reflect the diverse student body at LAPU, specific demographic data was not collected due to the anonymous nature of the EOC survey. This lack of demographic data may limit the ability to generalize the findings to the entire student population.\u003c/p\u003e\n\u003ch2\u003eProcedure\u003c/h2\u003e\n\u003cp\u003eWe analyzed the data using both a sentiment and thematic analysis. The purpose of using these two methods was to gain a deeper and more holistic understanding of the reasons why some students did not utilize the AI course assistant in their classroom. This dual methodology was designed to provide a robust and nuanced understanding of the data collected from open-ended survey responses.\u003c/p\u003e\n\u003ch2\u003eSentiment Analysis\u003c/h2\u003e\n\u003cp\u003eSentiment analysis was utilized to gauge the overall emotional tone of the students\u0026apos; responses regarding their non-use of the AI course assistant. This technique allowed us to quantify the attitudes expressed in the text, categorizing them into positive, negative, or neutral sentiments. Applying sentiment analysis allowed us to identify the general mood and attitudes of the students towards the AI course assistant. This provided an initial layer of understanding, highlighting the prevalent emotional reactions thatcould indicate broader patterns of feelings or attitudes towards the use of artificial intelligence within an online classroom.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eThematic Analysis\u003c/h2\u003e\n\u003cp\u003eWe also chose to include a thematic analysis to explore the responses from a qualitative perspective. The thematic analysis gives us specific reasons and contextual factors behind the students\u0026apos; responses. This qualitative method involved systematically identifying, analyzing, and reporting patterns (themes) within the data. Thematic analysis enabled us to uncover the underlying reasons for the students\u0026apos; non-usage of the AI course assistant by identifying recurring themes and sub-themes in their responses. This approach provided rich, detailed insights into the various factors influencing student behavior, including potential barriers, misconceptions, and areas for improvement in communications about the AI course assistant..\u003c/p\u003e\n\u003cp\u003eBy combining sentiment analysis with thematic analysis, we were able to achieve a comprehensive understanding of the students\u0026apos; experiences and perspectives. Sentiment analysis offered a broad overview of the emotional landscape, while thematic analysis provided depth and context to the specific issues raised by the students. This dual approach ensured that our study captured both the quantitative and qualitative dimensions of the data, leading to a more informed and actionable set of findings.\u003c/p\u003e\n\u003ch2\u003eData Collection\u003c/h2\u003e\n\u003cp\u003eThe data for this study was obtained by adding two questions to the\u0026nbsp;End-of-Course (EOC) survey. The added questions about the non-use of Spark were designed to identify if a student used Spark and gather free writing responses from students who did not use Spark. Question 40 was a multiple choice question which asked, \u0026ldquo;Did you use Spark, the AI Course Assistant, while participating in your course?\u0026rdquo;. If the respondent selected \u0026ldquo;no\u0026rdquo; to Question 40, Question 42 became available to them: \u0026ldquo;What is the primary reason(s) you did not use Spark within your course?\u0026rdquo;.\u003c/p\u003e\n\u003ch2\u003eData Preparation\u003c/h2\u003e\n\u003cp\u003eThe survey responses were compiled into an Excel spreadsheet. The relevant data for this analysis was located in Column B of the spreadsheet, corresponding to \u0026quot;Question 41.\u0026rdquo; Any responses that were missing or null were excluded from the analysis to ensure the accuracy and relevance of the sentiment analysis.\u003c/p\u003e\n\u003ch2\u003eSentiment Analysis\u003c/h2\u003e\n\u003cp\u003eTo gain a quantitative understanding of the emotional tone of the students\u0026apos; responses regarding their non-use of the AI course assistant \u0026quot;Spark,\u0026quot; we conducted a sentiment analysis using the TextBlob library. This analysis assigned a polarity score to each response, ranging from -1 (negative sentiment) to 1 (positive sentiment), with 0 indicating neutral sentiment.\u003c/p\u003e\n\u003ch3\u003eData Processing\u003c/h3\u003e\n\u003cp\u003ePrior to analysis, responses underwent preprocessing to remove punctuation, convert text to lowercase, and exclude common stop words. These words are typically articles (a, an, the\u0026hellip;), prepositions (in, on, at\u0026hellip;), conjunctions (and, or, but\u0026hellip;) and pronouns (I, you, he. she\u0026hellip;). Responses lacking content or deemed irrelevant were excluded to ensure a focused analysis.\u003c/p\u003e\n\u003ch3\u003eStatistical Analysis and Visualization\u003c/h3\u003e\n\u003cp\u003eDescriptive statistics summarized the sentiment distribution. A histogram was employed to show the frequency of sentiment categories, while a pie chart illustrated the proportion of positive, neutral, and negative sentiments. Additionally, a word cloud visualized frequently occurring terms to highlight common themes.\u003c/p\u003e\n\u003ch3\u003eLimitations and Biases\u003c/h3\u003e\n\u003cp\u003eThe sentiment analysis using TextBlob is subject to certain limitations, including its inability to capture sarcasm and nuanced context. Misclassification is possible when responses contain ambiguous language. Furthermore, the voluntary nature of survey participation introduces potential selection bias, and the absence of demographic data limits the generalizability of findings. The reliance on self-reported data introduces the possibility of response bias, where participants may not accurately or honestly report their true experiences or feelings, influenced by factors such as memory recall issues or social desirability bias (Podsakoff et al., 2003). These limitations highlight the need for cautious interpretation of the results and suggest areas for future research to address these potential biases.\u0026quot;\u003c/p\u003e\n\u003ch3\u003eInstruments, Tools, and Software\u003c/h3\u003e\n\u003cp\u003eThe analysis was conducted using Python, with the Pandas, TextBlob, Matlotlib, and WordCloud libraries. These libraries were chosen for their robustness and ease of use in handling text data and generating meaningful visualizations.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eThematic Analysis\u003c/h2\u003e\n\u003ch3\u003eData Organization\u0026nbsp;\u003c/h3\u003e\n\u003cp\u003eComments from column \u0026quot;Question 41\u0026quot; were systematically extracted and compiled into a list for detailed analysis. This step ensured that the data was ready for the coding process.\u003c/p\u003e\n\u003ch3\u003eFamiliarization with Data\u003c/h3\u003e\n\u003cp\u003eTo gain a comprehensive understanding of the content and context, the comments were read multiple times. This iterative reading facilitated the identification of initial patterns and emergent themes within the data.\u003c/p\u003e\n\u003ch3\u003eGeneration of Initial Codes\u003c/h3\u003e\n\u003cp\u003eEach comment was assigned an initial code, which acted as a descriptive label summarizing the core reason provided by the student for not using the AI course assistant. These codes were preliminary and subject to further refinement.\u003c/p\u003e\n\u003ch3\u003eTheme Identification\u003c/h3\u003e\n\u003cp\u003eThe initial codes were analyzed to identify broader themes that encapsulated the reasons for non-use. This involved grouping similar codes together to form overarching themes such as \u0026quot;Perception of Unnecessity,\u0026quot; \u0026quot;Lack of Interest or Motivation,\u0026quot; \u0026quot;External Issues,\u0026quot; \u0026quot;Forgetfulness or Awareness Issues,\u0026quot; \u0026quot;Lack of Familiarity with AI Tools,\u0026quot; \u0026quot;Technical or Usability Issues,\u0026quot; and \u0026quot;Preference for Traditional Methods.\u0026quot;\u003c/p\u003e\n\u003ch3\u003eTheme Review and Refinement\u003c/h3\u003e\n\u003cp\u003eThe identified themes underwent a rigorous review process to ensure they accurately represented the data. This involved evaluating the coherence and distinctiveness of each theme and making necessary adjustments to refine their definitions and boundaries.\u003c/p\u003e\n\u003ch3\u003eRe-categorization of Comments\u003c/h3\u003e\n\u003cp\u003eComments were re-categorized according to the refined themes to maintain consistency and accuracy. Any comments that did not fit within the predefined themes were categorized under \u0026quot;Other\u0026quot; to capture additional nuances.\u003c/p\u003e\n\u003ch3\u003eDefinition and Naming of Themes\u003c/h3\u003e\n\u003cp\u003eEach theme was clearly defined and appropriately named to reflect the underlying reasons for non-use. Representative comments were selected to illustrate the key points within each theme, providing a rich description of the data.\u003c/p\u003e\n\u003ch2\u003eReport Production\u003c/h2\u003e\n\u003ch3\u003eFrequency Analysis\u003c/h3\u003e\n\u003cp\u003eThe frequency of each theme was calculated to quantify the prevalence of different reasons among the students. This quantitative aspect complemented the qualitative analysis, offering a clearer picture of the dominant themes.\u003c/p\u003e\n\u003ch2\u003eVisualization\u003c/h2\u003e\n\u003cp\u003eA bar chart was generated to visualize the frequency distribution of the identified themes. This graphical representation helped highlight the most common reasons for not using the AI course assistant.\u003c/p\u003e\n\u003cp\u003eBy adhering to this structured methodology, the thematic analysis yielded a robust understanding of the reasons behind students\u0026apos; non-use of the AI course assistant, offering valuable insights for improving its adoption and effectiveness.\u003c/p\u003e"},{"header":"Results","content":"\u003ch2\u003eSentiment analysis\u003c/h2\u003e\n\u003cp\u003eThe sentiment analysis of the survey responses to Question 41, \u0026ldquo;What is the primary reason(s) you did not use Spark within your course?\u0026rdquo; revealed the following distribution:\u003c/p\u003e\n\u003cul class=\"decimal_type\"\u003e\n \u003cli\u003eNeutral Sentiment:\u0026nbsp;The majority of the responses (72.3%) were neutral, with a sentiment score of 0.00. Examples of neutral responses include statements like \u0026quot;I had no reason to use it\u0026quot; and \u0026quot;I did not need to use it.\u0026quot;\u003c/li\u003e\n \u003cli\u003ePositive Sentiment:\u0026nbsp;A smaller portion of the responses (27.7%) were positive, with sentiment scores greater than 0. Examples of positive responses include \u0026quot;Did not need to. I was able to do research on my own,\u0026quot; which had a sentiment score of 0.55.\u003c/li\u003e\n \u003cli\u003eNegative Sentiment:\u0026nbsp;There were no responses with negative sentiment scores, indicating a lack of strong negative feelings towards the tool.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3\u003eVisualizations\u003c/h3\u003e\n\u003cp\u003eThe histogram (see Figure 1) shows the distribution of sentiment scores across the responses. Most scores were clustered around 0, confirming the predominance of neutral responses. Positive sentiment scores were present but less frequent, while negative sentiment scores were absent.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe pie chart in Figure 2 illustrates the proportion of responses in each sentiment category. The distribution showed that neutral responses constituted the majority (72.3%), followed by positive responses (27.7%), with no negative responses.\u003c/p\u003e\n\u003cp\u003eThe word cloud (see Figure 3) visualized the most frequently occurring words and phrases from the survey responses. Common words included \u0026quot;need,\u0026quot; \u0026quot;reason,\u0026quot; and \u0026quot;use,\u0026quot; reflecting the main themes in the reasons provided by respondents.\u003c/p\u003e\n\u003ch3\u003eInterpretation\u003c/h3\u003e\n\u003cp\u003eThe sentiment analysis results indicate that most respondents did not have strong feelings, either positive or negative, about not using the tool. The predominance of neutral sentiment suggests that their reasons were more practical or situational rather than based on strong emotional responses.\u003c/p\u003e\n\u003cp\u003ePositive sentiments, though less frequent, showed confidence in alternative methods or tools. Respondents expressed satisfaction with their ability to perform tasks without the tool, indicating that their needs were met through other means.\u003c/p\u003e\n\u003cp\u003eThe absence of negative sentiments suggests that there is no widespread dissatisfaction with the tool itself. Instead, the reasons for not using the tool seem to stem from a perceived lack of necessity rather than any inherent issues with the tool.\u003c/p\u003e\n\u003ch3\u003eSummary of Key Sentiment Analysis Findings\u003c/h3\u003e\n\u003cul\u003e\n \u003cli\u003eNeutral Dominance:\u0026nbsp;The majority of responses were neutral, suggesting a lack of strong emotional responses towards not using the tool.\u003c/li\u003e\n \u003cli\u003ePositive Indications: A smaller portion of positive responses indicated satisfaction with alternative methods.\u003c/li\u003e\n \u003cli\u003eLack of Negative Feedback:\u0026nbsp;The absence of negative sentiments implies that the tool is not perceived negatively, but rather as unnecessary for certain users.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eThese insights suggest that further investigation into the alternative methods used by respondents and highlighting the unique benefits of the tool could help in increasing its adoption. Additionally, educational efforts to demonstrate the tool\u0026apos;s value might change the perception of those who currently see no reason to use it.\u003c/p\u003e\n\u003ch2\u003eThematic analysis\u003c/h2\u003e\n\u003cp\u003eThe thematic analysis resulted in comments fitting into 10 themes. Table 1 shows the themes, number of comments within the theme and three example statements from that theme.\u003c/p\u003e\n\u003cp\u003eTable 1. Themes, Number of Responses, Examples\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"624\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eTheme\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003eNumber of Responses\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eExamples\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003ePerception of Unnecessity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003e252\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eI did not need to use it.\u003c/p\u003e\n \u003cp\u003eI already knew where to go.\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eI did not feel the need to use it.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eLack of Interest or Motivation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003e41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eDid not want to.\u003c/p\u003e\n \u003cp\u003eNot interested.\u003c/p\u003e\n \u003cp\u003eHaven\u0026rsquo;t been interested in it.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eExternal Issues\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eI am too nervous to use it.\u003c/p\u003e\n \u003cp\u003eLack of time.\u003c/p\u003e\n \u003cp\u003eI didn\u0026rsquo;t get a chance.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eForgetfulness or Awareness\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003e42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eI had forgotten about spark.\u003c/p\u003e\n \u003cp\u003eI\u0026apos;m not sure what Spark is.\u003c/p\u003e\n \u003cp\u003eI am not familiar with it.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eLack of Familiarity with AI Tools\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003e37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eI\u0026apos;m not used to it.\u003c/p\u003e\n \u003cp\u003eI was not familiar.\u003c/p\u003e\n \u003cp\u003eNot familiar with it.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eTechnical or Usability\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003e30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eNot sure how to use.\u003c/p\u003e\n \u003cp\u003eIt was complicated.\u003c/p\u003e\n \u003cp\u003eDidn\u0026apos;t fully understand how to use it or its purpose\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003ePreference for Traditional Methods\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003e28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eI have my routine.\u003c/p\u003e\n \u003cp\u003eUsed textbook and other resources.\u003c/p\u003e\n \u003cp\u003eI don\u0026apos;t want to use AI.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eUsed Other Tools\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eUsed other resources.\u003c/p\u003e\n \u003cp\u003eI used ChatGPT.\u003c/p\u003e\n \u003cp\u003eI use a different AI tool\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eNot Helpful\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eI tried and didn\u0026apos;t like it. I prefer google, or bin\u003c/p\u003e\n \u003cp\u003eI don\u0026apos;t have time for the question/answer part of Spark. I am also very nervous about citing Spark.g AI copilot.\u003c/p\u003e\n \u003cp\u003eDo not care for it.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003eOther\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 110px;\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 306px;\"\u003e\n \u003cp\u003eNot sure.\u003c/p\u003e\n \u003cp\u003eNo reason, but will use it moving forward.\u003c/p\u003e\n \u003cp\u003eNo specific reason. I am still learning my way around the resources\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003ePerception of Unnecessity, Lack of Interest, and Lack of Awareness were the top three specific themes of why students did not utilize the AI course assistant other than the Other theme. See Figure 4 for the comparison of the amount of statements in each theme.\u003c/p\u003e\n\u003cp\u003ePerception of Unnecessity was the largest theme with 252 responses. The most common reason students did not use the AI course assistant is that they did not feel it was necessary. Many students believed the resources provided by their instructors were sufficient or that they could manage their coursework without additional assistance. Typical comments include,:\u0026rdquo; I did not need to use it,\u0026rdquo; \u0026ldquo;I already knew where to go,\u0026rdquo; and \u0026ldquo;I did not feel the need to use it.\u0026rdquo;\u003c/p\u003e\n\u003cp\u003eThere were 41 comments that fit the category of \u0026ldquo;Lack of Interest.\u0026rdquo; A significant number of students were simply not interested or motivated to use the AI assistant. This lack of engagement could be due to various factors, including personal preference or skepticism about the tool\u0026apos;s effectiveness. Typical comments that fit this category were, \u0026ldquo;Did not want to,\u0026rdquo; \u0026ldquo;Not interested,\u0026rdquo; and \u0026ldquo;Haven\u0026rsquo;t been interested in it.\u0026rdquo;\u003c/p\u003e\n\u003cp\u003eThere were 42 comments related to the theme of \u0026quot;Forgetfulness\u0026quot; and \u0026quot;Awareness.\u0026quot; In other words, some students forgot about the AI assistant or were not aware of its existence. Typical comments that fit this category were, \u0026ldquo;I had forgotten about Spark,\u0026rdquo; \u0026ldquo;I\u0026apos;m not sure what Spark is,\u0026rdquo; \u0026ldquo;I am not familiar with it.\u0026rdquo;\u003c/p\u003e\n\u003cp\u003eThe thematic analysis of student comments regarding their non-use of the AI course assistant revealed several significant insights into these three categories. \u0026nbsp;The most prevalent theme was a Perception of Unnecessity, with many students expressing that they did not find the AI assistant necessary for their coursework. This suggests that the resources, other than Spark, provided by instructors were perceived as sufficient or that students were confident in their ability to manage without additional help. This finding also highlights that the students did not understand that using the AI course assistants can help in learning \u0026nbsp;the content more quickly and making the information more durable (Reuell, 2019).\u003c/p\u003e\n\u003cp\u003eA notable number of students also cited a Lack of Interest or Motivation in using the AI assistant. This disinterest may stem from a lack of perceived value or skepticism about the tool\u0026rsquo;s effectiveness. External Issues such as time constraints and technical difficulties further hindered adoption, indicating that usability and compatibility are critical areas for improvement. This connects to and supports the findings of Al-Abdullatif, 2023 and Grassini, 2023 where they found that a students perception and familiarity with technological tools play a direct role in the students confidence to use the tool and willingness to use the tool.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eForgetfulness or Awareness Issues \u0026nbsp;were also identified, with some students either forgetting about the AI assistant or not being aware of its existence. This suggests that the implementation of AI tools in educational settings may require more than just an initial introduction. It highlights the need for enhanced communication strategies that go beyond a single announcement. Regular reminders through multiple channels, such as email, learning management systems, and in-class announcements, could ensure that students remain aware of the tools available to them.\u003c/p\u003e\n\u003cp\u003eForgetfulness may also stem from cognitive overload or competing priorities, especially in a busy academic environment. To mitigate this, integrating the AI assistant more seamlessly into the daily routines and course activities of students could make it a more natural part of their learning experience. For example, instructors could design assignments or in-class exercises that explicitly require the use of the AI assistant, thereby reinforcing its value and ensuring students become more familiar with its functions.\u003c/p\u003e\n\u003cp\u003eTargeted communication could help address awareness issues. For instance, emphasizing the benefits of using the AI assistant in terms of academic performance and time management may encourage more students to explore and use the tool consistently. Personalized messages that demonstrate how the AI assistant can be tailored to meet individual learning needs may also be effective in driving engagement (Rogers, 2003).\u003c/p\u003e\n\u003cp\u003eThese strategies align with broader principles of change management in education, where continuous engagement and clear messaging are critical for the successful adoption of new technologies (Rogers, 2003). Further research could explore the specific factors that contribute to forgetfulness or lack of awareness and assess the effectiveness of various communication strategies in improving the consistent use of AI tools in educational settings.\u003c/p\u003e\n\u003cp\u003eLack of Familiarity with AI Tools was another significant barrier, as some students expressed hesitation to use the AI assistant due to limited prior experience with such technology or discomfort with learning new tools. This issue highlights the need for targeted interventions to bridge the gap between students and the effective use of AI in their educational journey.\u003c/p\u003e\n\u003cp\u003eOne reason for this hesitancy may be the general anxiety associated with adopting new technologies, especially among students who may already feel overwhelmed by the demands of their coursework. Jia and Tu suggest that familiarity breeds confidence, and providing students with early and frequent opportunities to engage with AI tools in a low-pressure environment could reduce their apprehension (2023). For instance, offering optional workshops, tutorials, or orientation sessions at the beginning of the term could introduce students to the AI assistant in a structured and supportive way.\u003c/p\u003e\n\u003cp\u003eOngoing support is crucial for the continued use and adoption of the AI tools. Instructors and support staff could offer regular check-ins or office hours dedicated to helping students troubleshoot any issues they encounter with the AI tools. Peer mentoring programs could also be an effective strategy, where more tech-savvy students assist their peers in learning and using the AI assistant. This approach not only fosters a collaborative learning environment but also encourages the sharing of best practices among students.\u003c/p\u003e\n\u003cp\u003eIntegrating AI tools into course activities and assignments gradually, rather than introducing them all at once, could help students build their proficiency over time. For example, starting with simple tasks that require the use of the AI assistant and gradually increasing complexity can make the learning curve less steep.\u003c/p\u003e\n\u003cp\u003eA small group of students preferred Traditional Methods of studying, relying on established routines or other recommended programs. This preference indicates that the AI assistant needs to clearly demonstrate its unique benefits to convince students of its added value.\u003c/p\u003e\n\u003ch3\u003eSummary of Results\u003c/h3\u003e\n\u003cp\u003eThe sentiment and thematic analyses together reveal that the primary barriers to using the AI course assistant are practical in nature rather than rooted in strong negative sentiments. The predominance of neutral responses underscores a perception of unnecessity, a lack of interest, and forgetfulness or unawareness as key factors. The sentiment and thematic analyses of student responses to the question regarding their non-use of the AI course assistant, Spark, reveal several key insights.\u003c/p\u003e\n\u003ch3\u003eSentiment Analysis Findings\u003c/h3\u003e\n\u003cp\u003eNeutral Sentiment Dominance: The majority of responses (72.3%) exhibited neutral sentiment, indicating a lack of strong emotional responses towards not using the tool. Students commonly expressed practical or situational reasons, such as \u0026quot;I had no reason to use it.\u0026quot;\u003c/p\u003e\n\u003cp\u003ePositive Sentiment: A smaller portion of responses (27.7%) were positive, reflecting confidence in alternative methods or satisfaction with other tools. Comments such as \u0026quot;I was able to do research on my own\u0026quot; illustrate this sentiment.\u003c/p\u003e\n\u003cp\u003eAbsence of Negative Sentiment:\u0026nbsp;No negative sentiment was detected, suggesting that the tool is not viewed unfavorably, but rather as unnecessary for some students.\u003c/p\u003e\n\u003ch3\u003eThematic Analysis Findings\u003c/h3\u003e\n\u003cp\u003ePerception of Unnecessity: The most common theme, with 252 responses, was the belief that the AI assistant was not needed. Students felt that the resources provided by their instructors were sufficient or that they could manage their coursework without additional help.\u003c/p\u003e\n\u003cp\u003eLack of Interest or Motivation: A significant number of students (41 responses) expressed disinterest in using the AI assistant, often due to personal preference or skepticism about its effectiveness.\u003c/p\u003e\n\u003cp\u003eForgetfulness or Awareness Issues: Some students (42 responses) forgot about the AI assistant or were unaware of its existence, highlighting the need for improved communication and regular reminders.\u003c/p\u003e\n\u003cp\u003eLack of Familiarity with AI Tools: Another barrier identified was students\u0026apos; discomfort with new technology, underscoring the importance of training and support to help students become more comfortable with AI tools.\u003c/p\u003e\n\u003cp\u003eExternal Issues and Traditional Methods: Other factors such as external pressures, technical difficulties, and a preference for traditional study methods also contributed to the non-use of the AI assistant.\u003c/p\u003e\n\u003cp\u003eThe results indicate that most students did not perceive a strong need for the AI assistant, likely due to the adequacy of existing resources or confidence in their own abilities. The lack of negative sentiment suggests that the tool itself is not seen as problematic, but rather that its perceived necessity and value may need to be communicated more effectively. Addressing issues of awareness, familiarity, and support will be crucial in increasing the adoption of AI tools in educational settings.\u003c/p\u003e\n\u003cp\u003eThese findings suggest that with the right interventions, the AI assistant has the potential to be a valuable resource for more students, particularly if efforts are made to address the barriers identified in this study.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe findings from the sentiment and thematic analyses provide significant insights into students\u0026apos; perceptions and use of AI course assistants like Spark. The predominant theme of perceived unnecessity suggests that students generally do not view these tools as essential for their academic success. This perception aligns with the sentiment analysis, where a large proportion of neutral responses indicated a practical rather than emotional rationale for non-use.\u003c/p\u003e\n\u003cp\u003eIn both of these cases, it is apparent that students are not aware of the value of using tools such as Spark to engage in active learning. Students learn more with active learning, yet they enjoy the process less and believe they learn more with traditional lectures (Reuell, 2019). In light of this data, we are more able to focus our efforts towards specific areas to help students engage with Spark and achieve greater learning through engaging in active learning with Spark. This suggests a need for better communication and reminders about the availability and benefits of using the tool.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003ePracticality Over Necessity\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe overwhelming sentiment of neutrality indicates that many students do not perceive the AI course assistant as a necessary tool in their learning process. This suggests a possible gap between the potential benefits of AI course assistants and students\u0026apos; understanding or experience of these benefits. As Reuell (2019) indicated, active learning environments enhance learning retention and outcomes, yet students may not see how AI tools like Spark contribute to this active learning process. The findings indicate a need for better communication of how these tools can enhance learning efficiency and support academic success.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eLack of Engagement and Awareness\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe theme of Lack of Interest or Motivation suggests that some students are not compelled to use AI assistants due to a lack of perceived value or skepticism about their effectiveness. Moreover, the themes of \u0026quot;Forgetfulness or Awareness\u0026quot; and \u0026quot;Lack of Familiarity with AI Tools\u0026quot; highlight the importance of increasing awareness and understanding of AI capabilities. Many students indicated forgetfulness or unfamiliarity with Spark, suggesting that enhanced communication and training could address these barriers.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eTechnical and Usability Concerns\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe analysis also pointed out technical or usability issues as a barrier to adoption. While not the most prominent reason for non-use, the existence of these concerns emphasizes the need for continuous improvement of AI tools to ensure the tools are user-friendly and compatible with various student needs and technological proficiencies. This aligns with the need for training and support systems that can help bridge the gap between students and technology, making the transition smoother and more accessible.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eTraditional Methods and Resistance to Change\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe preference for traditional methods, as noted in the thematic analysis, underscores a resistance to change that is common in educational settings. Students\u0026apos; reliance on established routines and resources suggests that AI tools need to demonstrate clear and unique benefits to be considered valuable additions rather than mere alternatives. It is crucial to highlight these features of AI course assistants that distinguish them from traditional methods, potentially increasing their attractiveness and adoption among students.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003eRecommendations\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eBased on these findings, several strategies are recommended to increase the adoption and effectiveness of AI course assistants:\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003e\u003cstrong\u003eEnhanced Communication and Training:\u003c/strong\u003e Institutions should focus on raising awareness for both students and faculty \u0026nbsp;about AI tools through informational sessions, demonstrations, and success stories to illustrate their benefits. Providing training sessions to familiarize students and faculty with the technology can alleviate hesitations related to unfamiliarity.l This would come in the form of a series of short videos (less than 90 seconds) that demonstrate capabilities and create awareness. Videos would be delivered via emai, course pop-ups, and within the universities app.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eHighlighting Unique Benefits:\u003c/strong\u003e Clearly communicate the specific advantages of AI course assistants over traditional methods, such as personalized learning paths and real-time feedback, to enhance learning experiences.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eEngaging Students:\u003c/strong\u003e Increase motivation by integrating AI tools into regular coursework and showcasing their effectiveness through practical examples and testimonials.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eProvide Training and Support\u003c/strong\u003e: Early and ongoing support, including workshops and peer mentoring, can help students become more comfortable and proficient with AI tools.\u003c/li\u003e\n\u003c/ol\u003e\n\u003ch2\u003e\u003cstrong\u003eFurther Research\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eFuture research should explore the long-term impact of AI assistants on student learning outcomes and the specific features that most effectively drive engagement and success. By doing so, institutions can fully realize the benefits of AI integration in education, ultimately supporting students in achieving their academic goals.\u003c/p\u003e\n\u003cp\u003eAnother important avenue for research is examining how professors\u0026apos; utilization and buy-in of AI course assistants in educational settings influence student use and engagement with these tools.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eAddressing these barriers is not merely about increasing tool adoption; it is about ensuring equitable access to transformative learning technologies that can bridge educational gaps, enhance lifelong learning, and prepare students for an increasingly AI-integrated society.\u0026nbsp;AI course assistants represent a promising avenue for enhancing educational experiences. By understanding and addressing the perceptions and barriers to their adoption, educational institutions can harness the full potential of these tools, ultimately creating a more dynamic and personalized learning environment. This will require a concerted effort to increase awareness among students and faculty, provide training, and continuously improve the usability of AI tools to ensure their successful integration into academic settings.\u003c/p\u003e\n\u003cp\u003eThis study explored the factors contributing to some students\u0026apos; non-utilization of the AI course assistant Spark by students at Los Angeles Pacific University, as well as strategies to enhance Spark\u0026rsquo;s utilizaton and engagement by students.. Through the analysis of student feedback, several key insights were revealed that address our research questions.\u003c/p\u003e\n\u003cp\u003eRQ1: \u0026quot;What are the primary factors (e.g., technological, psychological, educational) that contribute to some students not utilizing the AI course assistant Spark available in their online classroom?\u0026quot; The findings indicate that the predominant barrier is a perception of unnecessity, where students feel that existing resources are sufficient and do not perceive additional value in using the AI assistant. Lack of interest or motivation and forgetfulness or unawareness also emerged as significant factors, suggesting that many students are not fully aware of the assistant\u0026apos;s capabilities or benefits. Additionally, a preference for traditional study methods and technical usability issues were identified as barriers, highlighting the psychological and educational aspects influencing student decisions.\u003c/p\u003e\n\u003cp\u003eRQ2: \u0026quot;What strategies can be implemented to increase the utilization of the AI course assistant Spark among students, thereby enhancing their learning experience and academic performance?\u0026quot; The study suggests several actionable strategies. Enhancing communication and training about the AI assistant\u0026apos;s features and benefits can address issues of awareness and familiarity. Highlighting the unique benefits of the AI assistant, such as personalized learning and real-time feedback, can motivate students by demonstrating clear advantages over traditional methods. Engaging students through integration of the AI assistant into regular coursework and showcasing its effectiveness with practical examples can further encourage adoption.\u003c/p\u003e\n\u003cp\u003eWhile the AI course assistant Spark offers significant potential to enhance learning experiences, its adoption is hindered by perceptions of unnecessity and a lack of awareness. By addressing these barriers through targeted strategies, educational institutions can better leverage AI tools to create more dynamic and personalized learning environments.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eThe research was ethically approved by the Institutional Review Board (IRB) of Los Angeles Pacific University. Prior to participation, all participants were duly informed of their rights and responsibilities and provided explicit written consent. The study was conducted in agreement with the guidelines governing research involving human participants, as outlined by the IRB .\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eThe manuscript was a collaborative piece between G.H. and C.S. Both authors contributed to each section of the paper with one author being primarily responsible. G.H. and C.S. analyzed the data separately and as a team. Data gathering was a primary responsibility of G.H. G.H. was the primary writer of the data analysis and discussion sections. C.S. wrote the literature review. Both G.H. and C.S. wrote the introduction and conclusion sections. Both authors reviewed the manuscript.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eData is provided at this link https://docs.google.com/spreadsheets/d/1I9rkVPzO1L-C91rpS-Q2WLGJh3U0SjaJcMR82mSu-2Q/edit?usp=sharing\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAl-Abdullatif, A. M. (2023). Modeling students\u0026rsquo; perceptions of chatbots in learning: Integrating technology acceptance with the value-based adoption model. \u003cem\u003eEducation Sciences\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(11), 1151. https://doi.org/10.3390/educsci13111151\u003c/li\u003e\n\u003cli\u003eChen, X., Zou, D., Cheng, G., \u0026amp; Xie, H. (2022). AI in education: A review of current research and applications. \u003cem\u003eJournal of Educational Technology \u0026amp; Society, 25\u003c/em\u003e(1), 31-44.\u003c/li\u003e\n\u003cli\u003eDeng, X., \u0026amp; Yu, Z. (2023). A meta-analysis and systematic review of the effect of chatbot technology use in sustainable education. \u003cem\u003eSustainability\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(4), 2940. https://doi.org/10.3390/su15042940\u003c/li\u003e\n\u003cli\u003eEssel, H. B., Vlachopoulos, D., Tachie-Menson, A., Jhonson, E. E., Baah, P. K. (2022). The impact of a virtual teaching assistant (chatbot) on students\u0026apos; learning in Ghanaian higher education. \u003cem\u003eInternational Journal of Educational Technology in Higher Education\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(1), 57. https://doi.org/10.1186/s41239-022-00362-6\u003c/li\u003e\n\u003cli\u003eGrassini, S. (2023). Shaping the future of education: Exploring the potential and consequences of AI and ChatGPT in educational settings. \u003cem\u003eEducation Sciences\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(7), 692. https://doi.org/10.3390/educsci13070692\u003c/li\u003e\n\u003cli\u003eHanshaw, G., \u0026amp; Miller, K. (2024). Evaluating the impact of real-time AI feedback on student writing: Randomized control trial. Manuscript in preparation. Los Angeles Pacific University.\u003c/li\u003e\n\u003cli\u003eIlieva, G., Yankova, T., Klisarova-Belcheva, S., Dimitrov, A., Bratkov, M., \u0026amp; Angelov, D. (2023). Effects of generative chatbots in higher education. \u003cem\u003eInformation\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e(9), 492. https://doi.org/10.3390/info14090492\u003c/li\u003e\n\u003cli\u003eLabadze, L., Grigolia, M., \u0026amp; Machaidze, L. (2023). Role of AI chatbots in education: Systematic literature review. \u003cem\u003eInternational Journal of Educational Technology in Higher Education\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e(1), 1-17. https://doi.org/10.1186/s41239-023-00426-1\u003c/li\u003e\n\u003cli\u003eLuckin, R., Holmes, W., Griffiths, M., \u0026amp; Forcier, L. B. (2016). Artificial intelligence in education: Promises and implications for teaching and learning. Learning, Media and Technology, 41(1), 1-19. https://doi.org/10.1080/17439884.2016.1139939\u003c/li\u003e\n\u003cli\u003ePodsakoff, P. M., MacKenzie, S. B., Lee, J. Y., \u0026amp; Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879-903. https://doi.org/10.1037/0021-9010.88.5.879\u003c/li\u003e\n\u003cli\u003eReuell, P. (2019, September 4). \u003cem\u003eStudy shows that students learn more when taking part in classrooms that employ active-learning strategies\u003c/em\u003e. Harvard Gazette. https://news.harvard.edu/gazette/story/2019/09/study-shows-that-students-learn-more-when-taking-part-in-classrooms-that-employ-active-learning-strategies/\u003c/li\u003e\n\u003cli\u003eRogers, E. M. (2003). \u003cem\u003eDiffusion of innovations\u003c/em\u003e (5th ed.). Free Press.\u003c/li\u003e\n\u003cli\u003eWilliams, R. T. (2024). The ethical implications of using generative chatbots in higher education. \u003cem\u003eFrontiers in Education\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e, 1331607. https://doi.org/10.3389/feduc.2023.1331607\u003c/li\u003e\n\u003cli\u003eWu, R., \u0026amp; Yu, Z. (2023). Do AI chatbots improve students\u0026rsquo; learning outcomes? Evidence from a meta‐analysis. \u003cem\u003eBritish Journal of Educational Technology\u003c/em\u003e, \u003cem\u003e55\u003c/em\u003e(1), 10-33. https://doi.org/10.1111/bjet.13334 \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"discover-artificial-intelligence","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"diai","sideBox":"Learn more about [Discover Artificial Intelligence](https://www.springer.com/44163)","snPcode":"","submissionUrl":"","title":"Discover Artificial Intelligence","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"ai course assistants, educational technology, student engagement, mixed-methods research, technology adoption, learning enhancement, barriers to technology use, sentiment analysis, thematic analysis, higher education","lastPublishedDoi":"10.21203/rs.3.rs-5867866/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5867866/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis research investigates the factors behind the non-utilization of the AI course assistant Spark among students at Los Angeles Pacific University (LAPU). Despite AI\u0026rsquo;s proven ability to enhance academic performance, motivation, and efficiency, a significant portion of students choose not to engage with this technology. Through a mixed-methods exploratory approach, key barriers to adoption are identified, including perceptions of unnecessity, lack of interest, and unfamiliarity with AI tools. External challenges, such as technical issues and a preference for traditional learning methods, are also examined. By combining sentiment analysis with thematic analysis of student survey responses, the findings offer a comprehensive understanding of the reasons for non-use. Targeted strategies\u0026mdash;improving communication about AI\u0026rsquo;s benefits, providing training to build familiarity, and integrating AI more seamlessly into coursework\u0026mdash;are recommended to increase adoption. Addressing these barriers is crucial, as non-utilization not only limits individual academic growth but also undermines efforts to equip students with essential skills for an AI-driven future. This research contributes to the growing literature on AI in education and provides actionable insights for educators and institutions seeking to maximize the impact of AI on learning outcomes.\u003c/p\u003e","manuscriptTitle":"Exploring Barriers to AI Course Assistant Adoption: A Mixed-Methods Study on Student Non-Utilization","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-02-10 10:24:29","doi":"10.21203/rs.3.rs-5867866/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-03-12T14:41:46+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-02-27T01:58:29+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-02-24T22:44:48+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-02-23T04:03:36+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"45174721456775451523900320335242206725","date":"2025-02-22T03:37:15+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-02-21T12:42:39+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"194094172438162026155005030847570892228","date":"2025-02-19T04:12:50+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"65211355290178485458988179310582678609","date":"2025-02-19T03:44:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"79870047484541493982694398614724885688","date":"2025-02-18T22:55:11+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"43444821839389953878473278265206147288","date":"2025-02-18T14:51:24+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-02-17T03:32:47+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-02-12T00:51:45+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-02-06T13:50:36+00:00","index":"","fulltext":""},{"type":"submitted","content":"Discover Artificial Intelligence","date":"2025-01-20T17:13:48+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"discover-artificial-intelligence","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"diai","sideBox":"Learn more about [Discover Artificial Intelligence](https://www.springer.com/44163)","snPcode":"","submissionUrl":"","title":"Discover Artificial Intelligence","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"47d72894-e067-47be-be43-2171bfa36c28","owner":[],"postedDate":"February 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-08-07T07:22:59+00:00","versionOfRecord":{"articleIdentity":"rs-5867866","link":"https://doi.org/10.1007/s44163-025-00312-x","journal":{"identity":"discover-artificial-intelligence","isVorOnly":false,"title":"Discover Artificial Intelligence"},"publishedOn":"2025-07-25 15:58:09","publishedOnDateReadable":"July 25th, 2025"},"versionCreatedAt":"2025-02-10 10:24:29","video":"","vorDoi":"10.1007/s44163-025-00312-x","vorDoiUrl":"https://doi.org/10.1007/s44163-025-00312-x","workflowStages":[]},"version":"v1","identity":"rs-5867866","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5867866","identity":"rs-5867866","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}