List of Figures

List of Abbreviations

AAC — Augmentative and alternative communication
AI — Artificial intelligence
ANN — Artificial neural networks
AR — Augmented reality
ASD — Autism spectrum disorder
CoDOA — Cognitive Development Optimization Algorithm
CTA — Cognitive Tutor Algebra
CTD — Constant Time Delay
CVT — Control-value theory
EI — Explicit instruction
EFM — Electronic fetal monitoring
FLSM —  Felder and Silverman’s Learning Style Model
HMD — Head-mounted display
ID — Intellectual disability
ICT — Information and communication technology
IeLS  — Intelligent eLearning System
ITS — Intelligent tutoring system
LD — Learning disability
LMS — Learning management system
ML — Machine learning
MOOC — Massive Open Online Course
NLP — Natural language processing
PCLS — Personalized creativity learning system
POV-VM — point-of-view video modeling
PVLE — Personalized Virtual Learning Environment
SRL — Self-regulated learning
STT — Speech-to-text
TTS — Text-to-speech
USD — United States Dollar
VLE  — Virtual Learning Environment
VR — Virtual reality

Executive Summary

Overview

This report explores how artificial intelligence-driven Education 4.0 can transform online learning by providing highly personalized, engaging, and adaptive educational experiences that meet the needs of individual students. With rapid advancements in AI and machine learning, tools for personalized learning, dynamic assessments, and immersive technologies like virtual reality and augmented reality have become increasingly accessible, primarily through learning management systems provided by companies such as CourseClout. By incorporating these tools, online learning can bridge educational gaps, increase engagement, and enhance flexibility for both learners and educators.

Key Findings

1. Advancements in AI and Machine Learning

  • Modern artificial intelligence (AI) capabilities, such as machine learning algorithms and intelligent tutoring systems (ITSs), allow for real-time adaptation of content and assessments.
  • Machine learning enables course creators to customize educational materials to individual students’ strengths, weaknesses, and learning styles, fostering a personalized learning path.

2. Personalized Learning and Engagement

  • AI-powered platforms can tailor educational experiences, enhancing students’ enjoyment, motivation, and engagement. This is crucial as emotions like boredom significantly impact academic performance, while positive emotions improve focus and flexibility
  • Interactive learning experiences, such as gamification, simulations, and immersive VR/AR environments, make abstract concepts tangible and support students’ ability to master complex content.

3. Enhance Accessibility for Disadvantaged Students

  • AI-based systems and features, including real-time language translation, customized content, and accessibility tools, make online learning more inclusive for students with disabilities and those in underserved regions.
  • By personalizing education pathways, AI can identify and support at-risk students, ensuring they receive appropriate resources and interventions to succeed.

4. Empowering Educators and Course Creators

  • AI assists educators by automating routine tasks such as grading and attendance tracking, allowing them to focus on high-value student interactions.
  • AI-driven video creation tools and content generation enhance educators’ flexibility, making high-quality, accessible learning materials readily available. Companies like CourseClout specialize in end-to-end AI-enhanced online course creation, supporting creators in delivering impactful, tailored educational experiences

CourseClout’s Role

CourseClout combines AI and high-production-quality video content to create engaging, adaptive online courses for course creators, businesses, and educators. Using generative AI, CourseClout provides customizable solutions, from interactive content to multi-language support, helping address global education challenges and making quality learning experiences accessible and enjoyable.

Conclusion

AI-driven online education, as exemplified by CourseClout, represents a promising solution to global learning disparities. Through personalized and adaptive learning pathways, dynamic assessments, and immersive content, AI addresses critical needs in education today. By making learning accessible and relevant, online education can complement efforts to tackle the global education crisis, increase engagement, and foster educational equity worldwide.

About the Author

Jacob J. P. Le Grange is a graduate of the esteemed KU Leuven. He currently serves as a research consultant for CourseClout. Jacob has published work on education and psychology in international journals and is keenly interested in using statistical simulations to inform psychological research and methodology. He resides in Brussels, Belgium.

https://orcid.org/0000-0002-5530-0284
https://www.linkedin.com/in/legrangejjp/
johanlegra@gmail.com

Transforming Learning with AI: CourseClout’s Insights into Education 4.0 and Online Learning

“AI can personalize learning pathways in the Metaverse, adapting to individual needs and styles.”

— Mark Zuckerberg (Entrepreneur and co-founder of Facebook)

“AI can analyze VR/AR learning data to provide insights and improve educational content.” —

Bill Gates (Business magnate, software developer, and philanthropist)

“AR has the potential to revolutionize the way we learn, bringing textbooks and historical sites to life right before our eyes.” —

Tim Cook (CEO of Apple Inc. and technology executive)

Chapter 1: Education 4.0

Education 4.0 marks a significant shift in educational practices, driven by globalization and rapid information and communication technology (ICT) advancements, particularly the rise of big data and artificial intelligence (AI; Awad et al., 2022; Rahimi & Oh, 2024). However, AI and big data are not new. It has been researched and applied to education for over 50 years (for a more extensive delineation, see Luan et al., 2020). For example, the first AI-powered intelligent tutoring system (ITS) was SCHOLAR. SCHOLAR was created to enhance geography instruction by producing interactive answers to students’ statements and questions about the learning material (Carbonell, 1970). Moreover, early research linking AI and education focussed on developing education chatbots, robotic systems, and intelligent tutoring systems (Luan et al., 2020).  However, the amount of data in the ITS systems was too little for data analyses (Luan et al., 2020), and back then, computers did not have sufficient computational power to harness the potential of AI.

Fast-forward a couple of decades, and technological advances in graphics processing units (GPUs),  tensor processing units (TPUs), and cloud computing bolstered computing power and opened up the age of AI (Klašnja‐Milićević et al., 2017; Krenn et al., 2022). In other words, we are now able to process large amounts of data faster and apply more complex models to data to understand the patterns in them, and allow machines to learn independently from us about the data (Klašnja‐Milićević et al., 2017; Krenn et al., 2022; Liu & Ardakani, 2022). Advancements in algorithms have been crucial in enabling AI systems to extract meaningful insights from large datasets (Munir et al., 2022).  Machine learning (ML) algorithms, a subset of AI, are designed to learn from data and improve their performance over time (Munir et al., 2022). ML algorithms are invaluable for online education. For example, an online course with many students may be hosted online on a learning management system (LMS), where students can interact, complete quizzes, review content, and apply their knowledge. I say many students since the average number of students in a typical Massive Open Online Course (MOOC) is 25,000 (Jordan, 2015).

 An advanced ML algorithm can analyze each student’s progress (e.g., identifying that the student has watched the first two videos) and provide tailored exercises to assess their understanding. Moreover, these algorithms can then use the progress to identify areas where the student struggles and adapt the learning experience to ensure that the student masters the content. For example, such AI algorithms may decide to adjust the pace, exercises, or even the content of the learning plan based on its evaluation of the student’s progress. Over time, such ML systems refine their approach and become more attuned to what is needed to help a specific student learn essential concepts in a course. Moreover, considering the substantial amount of data it receives from a large group of students, it becomes more finely tuned to the best course of action for different types of students. ML algorithms, therefore, learn from the behavioral data of each student and adapt intelligently to ensure that learning happens at the student’s pace.

Besides, ML algorithms trained on large amounts of visual data can now generate videos based on text input from scratch. Examples include Runway,  HeyGen, and Luma Dream Machine. This means that, if needed, an efficient AI-powered online course could create videos from scratch to explain complex content to students. Notably, CourseClout uses a variety of AI tools that speed up course creation while still allowing for quality video content. One such example is HeyGen, which not only creates AI avatars (e.g., people presenting a course) but also translates audio into more than 175 languages. We consider the various ways AI aids learning in later sections of this report.

In light of this brief and simple example, one can understand why these technological advances have led to a research explosion in the use of AI in Education since 2015 (Munir et al., 2022; Zawacki-Richter et al., 2019). Moreover, as quantified below, this explosion becomes understandable, given its promise for personalized education.

1.1. AI in Education: A Quick View of the Numbers

1.1.1 The Markets

We are experiencing an immense increase in global investment in ICT for education, with AI, ML, and big data arising as key focus areas for these investments (Li & Lalani, 2020). This upsurge follows the anticipation of global ICT investment proliferating due to the adoption of emerging technologies such as AI (SNS Insider Pvt. Ltd., 2023). In 2023, a report by SNS Insider estimated the global next-generation ICT market size at 35.7 billion USD and projected this market value to increase to 175.17 billion USD in 2032 (SNS Insider Pvt. Ltd., 2023). Massive Open Online Course (MOOC) platforms such as Coursera, EdX, and Udacity attract thousands of learners globally, representing an undeniable surge in online learning (Cooper & Sahami, 2013). With an increase in the number of students enrolled in MOOCs from 300,000 in 2011 to 220 million in 2020 (Shah, 2020), the popularity of these platforms inevitably signals significant investments in ICT infrastructure related to education. According to Papanastasiou et al. (2019), startups developing virtual reality (VR) and augmented reality (AR) systems experienced a 300% investment growth in 2016. At least, in part, this reflects the demand to use these devices in education (Inderjeet & Bhardwaj, 2024; Sanabria & Arámburo-Lizárraga, 2017). AI can be seen as an area of ICT that is currently thrown around as a buzzword everywhere, especially since the rise in popularity of chatbots like OpenAI’s ChatGPT and Google’s Gemini. Nevertheless, investments in using AI in education reflect a golden era for AI research in EdTech.

Global Market Insights Inc (2021, as cited in Majeed, 2023) projects that investments in using AI in the education market will soar to 20 billion USD by 2027, a substantial increase from an estimated 1 billion USD in 2020. Precedence Research (2024) also suggests that the global AI market will surge from 5.18 billion USD in 2024 to 112.30 billion USD in 2034.  At the very least, these numbers indicate that the markets strongly believe in AI’s potential for education and that investments are active in its research, development, and implementation in the classroom.

1.1.2. Scientific Research: Learner-Centered Models in Education with Technology

Above, I briefly mentioned the explosion in AI research in education over the past years. It is usually an excellent idea to look at review articles to understand the landscape of scientific inquiry surrounding a research topic. Review articles also offer a great way to access the existing evidence in a field. Review articles in scientific research provide the highest quality of research evidence, more so than individual studies (Spring & Hitchcock, 2009; Steglitz et al., 2015). For more information on the role of review articles in scientific inquiry, refer to the brief review of Le Grange and Van Den Noortgate (2024). Chapter 2 will discuss the significance of review articles for this report.

The literature on AI research in education has been increasing rapidly, given AI’s potential to personalize education for a large group of students —in other words, to personalize education cost-effectively and efficiently in a way that was impossible without technology (Almusaed et al., 2023; Liu & Yu, 2022; Luan et al., 2020; Munir et al., 2022). The prominent role of AI in EdTech research is now indisputable if one looks at some findings from research reviews. In their review, Belpaeme et al. (2018)  article examined 307 studies on the use of social robots (AI-powered) for education (1990-2016) and found that the number of studies increased significantly after 2010. Munir et al. (2022) analyzed 60 scientific journal articles on AI in digital education in another review article. They found that after 2015, the use of AI in education became more prominent as more researchers were attracted to conducting studies that explored its potential for education. Hinojo-Lucena et al. (2019) conducted a bibliometric study to analyze scientific research on using artificial intelligence in higher education, published in Web of Science (WoS) and Scopus databases from 2007 to 2017.

Interestingly, the authors found that only a relatively small number of scientific articles were published on the use of AI in higher education, thereby signaling that research on this topic is in its infancy. However, they found that a majority of publications on this topic were from published conference proceeding papers, which they suggest signals a high level of interest and discussion around the use of AI in higher education. While these review studies are not an exhaustive list of studies looking at the growth of AI-education research, they do illustrate an increased interest of the scientific community in finding ways to leverage and do experiments examining AI’s impact on learning and education. This increased interest is, at least in part, driven by a shift towards learner-centered models in education.

1.2 Leaner-Centered Models of Education

In education, learner-centered models prioritize the individual student’s needs, interests, and goals and contrast with traditional education, where teachers teach uniformly to a class despite individual differences in learning needs, interests, and goals (Ferguson, 2012; Kizilcec et al., 2017; Knowles et al., 2011; Kucirkova & Leaton Gray, 2023; Schmidt et al., 2009). There are various aspects to consider in learner-centered models:

Given the interest in these characteristics of learner-centered models of education, education research has researched and found ways in which educators can use AI to achieve personalized learning paths, allow for learner agency, and empower instructors to cater to different learning styles.

1.3. What is Education 4.0?

Education 4.0 is a new way of education that leverages the technologies of the Fourth Industrial Revolution, including AI, big data, and cloud computing, to create personalized, adaptive, and highly engaging learning experiences (Al-Badia et al., 2022; Bhutoria, 2022; Prinsloo et al., 2020; Yousafzai et al., 2016). Education 4.0 is, therefore, the efforts of educators, policymakers, companies, and researchers to use technology to truly a learner-centered education (see 1.2 Leaner-Centered Models of Education, above).

Technology now enables educators to provide a more personalized and engaging learning experience for each student, a capability that was previously impractical (Almusaed et al., 2023; Annuš, 2024; Gómez-Pulido et al., 2023; Luan et al., 2020). Personalized learning necessitates deeply understanding each student’s strengths, weaknesses, learning styles, and interests. Before technology, this would have required educators to manually assess and track individual student progress, which would have been incredibly time-consuming and demanding, especially in larger class sizes. Notably, traditional classrooms have high student-to-teacher ratios, making it difficult to carry out manual personalization (Costa et al., 2021).  Providing individualized instruction and support was difficult to scale without technology (Kucirkova & Leaton Gray, 2023). One-on-one tutoring, considered the ideal form of personalized learning, could have been more practical for most educational settings due to cost and resource constraints (Kucirkova & Leaton Gray, 2023).

However, the Internet and mobile computing have increased the amount of data that is now available about student learning (Gómez-Pulido et al., 2023; Klašnja‐Milićević et al., 2017). Any stakeholder involved in education can now collect more data than ever before at a fraction of the price and time it took two decades ago (Sahlberg, 2009). This data, whether captured through online learning platforms, digital assessments, or student interactions with technology, provides valuable insights into how students learn, their preferences, the behavior of their teachers, and even their socioeconomic status(Costa et al., 2021; Gómez-Pulido et al., 2023; Klašnja‐Milićević et al., 2017; X. Liu & Ardakani, 2022; Munir et al., 2022; Sønderlund et al., 2019). These massive chunks of data, in turn, help educators to identify weaknesses in students’ understanding or learning and adapt their instruction accordingly. Advancements in AI and machine learning techniques have made it possible to process and analyze these large quantities of data more efficiently and effectively for educators and made the job of personalized learning a lot easier(Akavova et al., 2023; Almusaed et al., 2023; Annuš, 2024; Majeed, 2023).

Unlike traditional education, Education 4.0 shifts the attention away from a one-size-fits-all approach to learning and instead relies heavily on technology to enhance student learning by personalizing the content and pace of education to each student´s needs, learning styles, strengths, and weaknesses (Al-Badia et al., 2022; Bhutoria, 2022; Palanisamy et al., 2021). In this way, Education 4.0 echoes the idea of personalized or precision medicine, where practitioners analyze large amounts of data to find patterns in individual patients’ health and use this to tailor their intervention and prevention efforts. With advances in AI, big data, cloud computing, and more, Education 4.0 can now realize the goal of creating personalized, adaptive, and engaging learning experiences that cater to individual student needs. In other words, the buzzword is personalization! Moreover, these benefits apply to students in online, face-to-face, or blended learning environments alike.

At this point in the report, it is helpful to identify and define some technologies that drive personalized learning —AI and big data.

1.3.1. Artificial Intelligence 

Stryker and Kavlakoglu (2024) define AI as “technology that enables computers and machines to simulate human learning, comprehension, problem solving, decision making, creativity and autonomy.” The literature reviewed in this report frequently describes AI as a field of computer science that aims to create systems capable of doing what humans do, including learning, problem-solving, reasoning, understanding natural language, recognizing patterns, and adapting to new information(Wartman & Combs, 2017; Zawacki-Richter et al., 2019).  AI systems rely heavily on data to execute these actions by learning from the data and adjusting its behavior over time (Gobert & Sao Pedro, 2017; Krenn et al., 2022). Moreover, it usually uses ML algorithms to identify patterns, make predictions, and optimize their responses based on the data they are trained on (Gobert & Sao Pedro, 2017; Inderjeet & Bhardwaj, 2024; Maghsudi et al., 2021). In Education 4.0, AI applications include intelligent tutoring systems, adaptive assessments, chatbots, and  VR and AR systems (Dhanayana et al., 2024; S. B. Vinay, 2023).

  • Intelligent tutoring systems (ITS) are precisely what their name suggests: systems that aim to simulate the guidance and support one would receive when going to a course tutor(Luckin & Holmes, 2016).
  • Adaptive assessments are a type of educational evaluation that utilizes AI algorithms to tailor the difficulty and content of the assessment in real-time based on the student’s performance (Akavova et al., 2023; Dhanayana et al., 2024). This dynamic adjustment ensures that students are appropriately challenged and that the assessment accurately reflects their current knowledge and skills(Akavova et al., 2023). The core of adaptive assessment relies on sophisticated AI algorithms that analyze student responses and dynamically adjust the assessment content (Dhanayana, 2024; Owoseni et al., 2024). These algorithms constantly evaluate student performance and select appropriate questions to match their proficiency level. It is important to note that these assessments mean that teachers do not manually do grading but that it is automatized (G. M. et al., 2024).
  • Chatbots: ChatGPT, which was only released for public use in 2022(Crompton & Burke, 2023), is an excellent example of a chatbot. Chatbots are computer programs that can converse with humans, typically through text-based interfaces (Annuš, 2024). They use artificial intelligence (AI), natural language processing (NLP), and machine learning (ML) techniques to understand and respond to user inputs (Almusaed et al., 2023; Outman et al., 2023). Chatbots promote active learning and increase student motivation by providing immediate feedback, personalized support, and an interactive learning environment (Almusaed et al., 2023). Moreover, they make educational support more accessible to all students, including those with disabilities or limited access to traditional support resources (Almusaed et al., 2023). Also, automating tasks and providing readily available support can lessen the burden of educators who constantly need to address students’ questions and concerns (Dwivedi et al., 2023; Kuhail et al., 2022).
  • VR/AR: VR is a computer-generated 3D virtual world that simulates the real world and enables users to interact with objects, concepts, and other individuals using motion sensors, head-mounted displays (HMDs), and tactile gloves (Martín-Gutiérrez et al., 2017; Orlosky et al., 2017). VR aims to place individuals in different (virtual) worlds, whereas AR changes aspects of the real world with virtual computer-generated images (Coursera Staff, 2024).
    • VR/AR technologies engage learners by providing a multisensory experience of the virtual world (Rui & Badarch, 2022) that includes a visual, auditory, and, more recently, a haptic experience of this world (Luan et al., 2020; Marougkas et al., 2023; Mikropoulos & Natsis, 2011). From my review of the recent Marougkas et al. (2023) review of VR/AR in education, the assertion of Mikropoulos and Natsis (2011) that the visual experience predominates the use of VR/AR technologies in education still holds. Nevertheless, adding the auditory experience to the virtual world provides a sense of immersion and interactivity that distinguishes VR/AR learning from learning in traditional classroom settings (see Papanastasiou et al., 2019). The crux of learning with VR/AR technology, according to Luan and colleagues (2020), is its ability to perceptually ground the learning experience by letting the student experience the content of a subject through more than one sensory modality. If embodied cognition theory (Barsalou, 2008) holds, Luan et al. (2020) posit that we may expect VR/AR technology to produce better learning outcomes than traditional education. Embodied cognition suggests that our mind, including our thoughts, language, emotions, and social interactions, are closely connected to our experience of our physical body and sensory experiences (Barsalou, 2008). Embodied cognition means that our understanding of the world is based on our sensory, motor, and introspective experiences, not just our thoughts (Barsalou, 2008). For example, when you encounter a dog, your concept of a dog is shaped by your previous sensory encounters with dogs. Rather than memorizing a detached, dictionary-like definition of a dog, your brain forms a simulation that encompasses the sensory experience of being with a dog: the visual impression of its fur, the sound of its bark, the tactile sensation of its fur when you touch it, and motor experience of petting it. So, instead of rote memorization of concepts, VR/AR technologies allow students to understand these concepts in the same way they learn about the world — i.e., via embodied cognition.
    • Educators can use VR/AR to break down complex, abstract concepts into interactive simulations (e.g., learning mathematical concepts: Cascales-Martínez et al., 2016). Students can choose their actions, learn through virtual games, and engage their senses, creating a more fun and interactive experience(Papanastasiou et al., 2019; M. Yang & Weng, 2023). These benefits hold not only for abstract concepts but also for virtual communal areas where students and online course attendees can collaborate on group projects (Almusaed et al., 2023). The possibilities are endless, and I will explore the applications of VR/AR in education in later sections of this report.

1.3.2. Big Data

Above, I stated how online platforms gather a large amount of data on students’ performance, learning behaviors, engagement, and so forth, as well as data on teachers. This large amount of data drives AI’s capacity to identify patterns, trends, and insights about individual learners (Klašnja‐Milićević et al., 2017). This capacity to identify trends at this level of detail is something that is highly improbable for even the most astute teachers in traditional classrooms and may not only help learning but also identify students who are at risk. We now live in an era where the internet and mobile technologies have allowed for the collection of vast chunks of individual data, and this is driving a revolution towards the realization of Education 4.0.

1.4. Technology-Practice Gap

All the benefits of AI for education I delineate in this report may seem impressive. However, there is an important caveat to consider: the technology-practice gap. This gap means that despite the increase in research examining the benefits of AI in education, its day-to-day application remains lagging (Rui & Badarch, 2022; Zawacki-Richter et al., 2019). Some authors also address AI’s use in terms of its potential as opposed to its current capabilities as used in traditional or online education (e.g., Almusaed et al., 2023; Murtaza et al., 2022; Owoseni et al., 2024; Yin, 2022). Nevertheless, several review articles (e.g., Almusaed et al., 2023; Costa et al., 2021; Gómez-Pulido et al., 2023; Luan et al., 2020; Marougkas et al., 2023) and primary research (e.g., Bin Rodzman et al., 2020; Carbonell, 1970; Fan et al., 2015; Inderjeet & Bhardwaj, 2024; Lynch & Ghergulescu, 2017; Pardamean et al., 2022; Soflano et al., 2015; Walkington & Bernacki, 2018) discuss the current use of AI in education, signaling progress.

Moreover, with COVID-19 forcing the world to adapt, we saw increased research towards using and adopting AI in education (Maghsudi et al., 2021). However, COVID is not the only reason for this. The increase in web-based learning systems, specifically in higher education, was also due to the higher number of students, limited teaching staff, and the development of new technologies (Maghsudi et al., 2021). Therefore, it is reasonable to conclude that while a gap exists between the potential of current AI technologies and their application in education, it is more nuanced than merely stating that AI’s implementation in education lags behind.

Chapter 2: Aims and Scope of The Report

2.1. Aims 

This report examines the role of AI in enhancing learning in online education. It aims to demonstrate how CourseClout’s services can use AI to produce engaging, personalized, and rewarding online courses.

CourseClout is a company that offers complete online course creation services for course creators, small and medium-sized businesses (SMBs), government education initiatives, academies, and enterprises. They provide end-to-end solutions for creating engaging, high-quality courses, including curriculum design, production, and LMS implementation. The company aims to enhance online learning, minimize dropouts, and boost learner engagement. They assist clients in creating and delivering effective educational content with striking videography to produce binge-able learning.

This report focuses on how course creators and educators can use technology in blended-learning or online learning environments. This report, therefore, focuses on:

  1. The realization of personalized learning in education is due to the implementation of AI technologies.
  2. How AI may aid educators and online course creators.

2.2. Scope of The Report

This report investigates how AI can drive personalized online learning by customizing content and assessments to adapt to an individual’s learning needs. Taking this perspective allows for exploring the impact AI can have on student engagement and performance, as well as facilitating online course creators and educators. Lastly, the report considers the ethical challenges of using AI in online education.

It remains crucial to reiterate what I discussed in the section on the technology-practice gap.

Implementing AI in education needs to catch up to its current technology capabilities. However, the implementation of AI technologies is computer-based, and, accordingly, most of its applications are in online settings or blended-learning environments as opposed to traditional face-to-face classroom settings (Munir et al., 2022; Popenici & Kerr, 2017; Zawacki-Richter et al., 2019). The increased accessibility of online education, that is, that you can attend them when and where you like, gives rise to more users and larger datasets than traditional education (Li & Dong, 2022; Murtaza et al., 2022). In addition, online teaching platforms and courses can adapt their content quicker and more effectively than conventional teaching methods (Zhu, 2022). Moreover, the use of AI chatbots to address real-time concerns of students (c.f., Almusaed et al., 2023; Owoseni et al., 2024) is only possible if traditional learning is supplemented with online learning (i.e., blended learning) or purely online settings. Also, as the needs of students and faculty progress, there is an increasing scope for incorporating digital education as a supplement to face-to-face education (Munir et al., 2022).

It should be noted that measuring the extent of the gap between AI technology and its current applications is complex and would require a more comprehensive systematic review of the literature. Although I recognize this gap, its full scope is beyond the purpose of this report, which focuses on offering a narrative review of the literature based on the aims mentioned above.

2.3. Process of Literature Review 

The literature for this report includes institutional reports, review articles, chapters in academic books, and data from institutional databases. The primary sources were academic databases such as ERIC and Web of Science, with Google Scholar occasionally used as a last resort. I also used secondary sources cited in the primary sources to expand the pool of relevant research. Additionally, Research Rabbit, an AI-powered tool, was employed to discover new academic research based on the existing pool of sources.

In the first paragraph of 1.1.2. Scientific Research: Learner-Centered Models in Education with Technology, I start highlighting the importance of review articles regarding scientific evidence. Given that this report can only provide a glimpse into the current landscape of AI research in education, I advise you to look at the review articles we use in this report. For more information, see References.

Chapter 3: The Rise of Personalized Learning Through AI

One of CourseClout’s goals is to create a tailored landing page and LMS for the courses they create. LMSs are software applications, in CourseClout’s case online platforms, that administer, document, track students’ progress, and deliver educational content (El-Bishouty et al., 2015; C. Li et al., 2023; Libbrecht et al., 2015). It is on these platforms that course enrollment happens, where course material (e.g., text, videos, powerpoints, and assessment) can be found, where students can contact each other or the instructor, and where assessments take place. Learning management systems have become an inextricable characteristic of learning management systems for over a decade (c.f., Martindale & Dowdy, 2010; Prinsloo et al., 2020).

At these LMSs, the benefits of AI can be developed, which explains why there is an increased movement towards making these systems AI-powered (Costa et al., 2021). I discuss below how AI use in LMS drives personalized education for educators and learners.  However, while most AI applications discussed below may happen via an LMS, I also consider additional ways AI may improve online education. For example, in the section on AI’s use in bolstering the capabilities of educators and online course creators, I also consider additional ways AI may transform online content creation.

3.1. Personalized Learning for Individual Learners 

3.1.1. Data? What Data?

As I have mentioned before, AI thrives as a result of data. So, any effective LMS needs to consider what data it needs to collect about students to ensure personalized learning. I first consider the data with, here and there, some brief mentions of its application. I consider the actual application and benefits of the data in the following sections.

One first point of data collection emerges when students enroll in a course. At enrollment, an LMS can gauge important demographic data such as age, gender, and prior education experience or subject knowledge (Liu & Yu, 2022). Since ML algorithms improve over time, as more students enroll and complete or drop out of an online course, these algorithms receive more data. As a result, the ML algorithm can more accurately utilize this information for prediction and become better with the predictions. Other information, such as prior academic records, is also collected by LMS platforms in higher education and provides an individualized learner profile (Annuš, 2024; Tempelaar et al., 2015). However, this is only a preliminary phase of data collection. Once enrolled, the actual data collection capabilities of LMSs are realized.

LMSs hosting an online course can access login and logout activities in amount, frequency, or time of day(Prinsloo et al., 2020). Once logged in, students interact with content and participate in various assessments. LMSs gather data on how learners engage with the content, including which content they access, the order of access, navigation between content and supplementary material, and time spent with the material. (Bhutoria, 2022; L. Sie et al., 2018; M. Liu & Yu, 2022; Prinsloo et al., 2020).

Importantly, interactions with learning materials are a way for LMSs to measure students’ learning styles(Bernard et al., 2016; El-Bishouty et al., 2015). Various learning styles exist, and the Felder and Silverman’s Learning Style Model (FSLSM) is a good introduction (Felder & Silverman, 1988). Course creators can find an example of this by using DeLeS as an add-on to a popular LMS called Moodle (El-Bishouty et al., 2015). DeLeS is a rule-based system designed to identify students’ learning styles based on their behavior within an LMS (El-Bishouty et al., 2015). Based on the type of content students access, the frequency in which they use them, the time spent on them, and how this translates to performance on quizzes, DeLeS can identify learning styles and even the working memory capacity of each student(El-Bishouty et al., 2015). Besides identifying learning styles, Some LMSs go so far as to allow educators and online course creators to measure when students are off-task, bored, or even frustrated during the learning process (R. S. J. D. Baker, 2007; D’Mello et al., 2008).

In many LMS systems, ITS is also a feature. Here, students can interact with chatbots to clarify their understanding of the content (Al-Badia et al., 2022; Bhutoria, 2022; Yildirim & Celepcikay, 2021). These interactions are a data gold mine insofar as they allow to gauge learner preferences, showcase how they solve problems, and help to identify common misconceptions in the learning plan (Al-Badia et al., 2022; Bhutoria, 2022; Yildirim & Celepcikay, 2021). Moreover, ITS data collection allows at least one way in which online course creators can ascertain data regarding students’ learning styles. I discuss these learning styles in subsequent sections of this chapter.

Another unique way learners can interact with content on LMSs is through VR/AR technologies. Here, the content relates to objects and scenarios in the virtual world and interactions with other students, providing myriad data on student learning (see Papanastasiou et al., 2019).

Communication and interaction also form an integral part of online course models. Students can interact on forums and chats, allowing course creators to understand how students integrate the course material (A. Dutt et al., 2017; Prinsloo et al., 2020). Communications, of course, is not only isolated to forums but permeates through VR/AR simulations. Moreover, students can access real-time chatbots to try and hone their understanding of a concept, which, in turn, provides course creators with an additional data source (Majeed, 2023).

Of course, besides engagement with teaching content and communication data, assessments also form an integral part of online learning by allowing students to identify their strengths, weaknesses, and mastery of the learning material. The time spent on questions, amount of retakes, answers, and overall test-taking behavior are vital sources for course creators about learners, how their courses are structured, and how to proceed with existent online courses(Almusaed et al., 2023; A. Dutt et al., 2017; M. Liu & Yu, 2022; Prinsloo et al., 2020).

Future suggestions for educational data include using cameras, such as computer cameras, to read facial expressions to ascertain engagement and emotions during learning (Chalmers, 2017; Dhananjaya et al., 2024; Rui & Badarch, 2022). Some sources, such as Giannakos et al. (2019), have accurately measured students learning by using physiological sensors (wristband measuring electrodermal activity, heart rate, blood volume), eye tracking technology, and an electroencephalogram (EEG). While the practical implementation of this may be costly and increase the education between developed and developing countries, these findings suggest a new era where technology becomes increasingly integrated into education. In time, it may become cheaper and more accessible to implement technologies such as these into online learning to gauge its effectiveness and to help course creators make learning more personalized.

3.1.2. The Many Faces of AI in Personalized Learning 

The plurality of data sources also means that there is a plurality of ways in which AI makes learning more personalized. The application I discuss is not an exhaustive list of ways AI improves learning by making it more personalized. Instead, it serves as a narrative overview of the current state of using AI in online education. In this report, I restrict the discussion of AI’s use in online education to (a) individualized learning paths, (b) feedback and support via dynamic assessments, (c) engagement and interactivity, and (d) the augmentation of learning material. I will only briefly discuss the augmentation of learning material as it pertains to the personalized education of students. However, a big part of the benefits of such augmentation is related to what it can do for those who are online course creators. Therefore, a bigger chunk will be discussed in the relevant section below.

3.1.2.1. Individualized Learning Paths

Learning styles and preferences take center stage in individualized learning paths (El-Bishouty et al., 2015; M. Liu & Yu, 2022; Owoseni et al., 2024). In this report, I define learning styles as the diverse ways in which students perceive, process, and retain the information of an online course (Clay & Orwig, 1999). The learning style should ideally drive the content used in online learning modules (Kumar & Ahuja, 2020).

In my discussion of individualized learning paths, I briefly consider how it relates to the pace and content of online learning. After that, I consider how course creators can use personalized learning paths to identify at-risk students and how this capability relates to our current understanding of the global education crisis. I will then connect this to our comprehension of self-directed learning and conclude by explaining how it can enhance inclusivity in online education. As I discuss these concepts, I will consider the theory and our knowledge of the anticipated and scientifically proven advantages of personalized learning paths.

In a personalized learning environment, the content can be readily adapted to match students’ current knowledge, interests, progress, and learning styles (U.S. Department of Education, 2017; Yousafzai et al., 2016). The idea is that AI-powered online courses monitor students’ progress throughout and, based on this, adjust the content based on each student’s mastery of the concepts (Akavova et al., 2023; Bhutoria, 2022). One way this can happen is to adjust the format, whether images, animations, or sounds, based on the sensorial preference of learners (Owoseni et al., 2024). Based on what works best, the ML algorithm may also decide whether students need more or less detail for each section of an online course (Owoseni et al., 2024). For example, Owoseni et al. (2024) suggest using large language models, such as ChatGPT, to make different versions of text students must read based on their reading level, prior knowledge, and course goals. Some students, based on their goals, may be more interested in some sections of an online course than others and, in such cases, an AI-drive online course can decide to include additional resources, ideas, and questions that relate to the sections of interest (Akavova et al., 2023; Dhananjaya et al., 2024; Owoseni et al., 2024). Another way the ML algorithms can adapt to an individual’s learning path relates to the student’s understanding of the course concepts understanding of concepts. For those who have mastered concepts, AI can gradually increase the difficulty of the content to keep challenging the learner (Akavova et al., 2023). In contrast, those who struggle may receive additional explanations or resources to help them master the content (Akavova et al., 2023). However, my discussion above on how AI can adapt online courses’ content remains theoretical.

Naturally, it is essential to consider the evidence of this in practice. In other words, does AI’s adaptation of learning content produce better learning outcomes in online learning environments? This report needs to consider answering this question; otherwise, the benefits of how AI adapts learning content based on individual learning characteristics remain mere theoretical conjecture. Please note, however, that some of the outcomes of studies discussed in this section pertain to more than the benefits of AI for learning due to its ability to provide personalized learning paths in online learning. Instead, some studies employ additional methods like dynamic assessment to produce better student results. For instance, in the section on dynamic assessment, the studies we discuss here that used dynamic assessment and content adaptation to achieve personalized learning also demonstrate how AI boosts online learning with dynamic assessments.

3.1.2.1.1. Empirical Evidence

Overall, in the single studies we review below, the AI adapted the content based on personalized learning profiles it created for each individual student. These studies suggest an increase in the learning motivation (Kose & Arslan, 2016; D. Xu & Wang, 2005), learning engagement (Walkington & Bernacki, 2018; D. Xu & Wang, 2005), better academic results for those who used an AI-powered learning platform as opposed to those who did not (Hwang et al., 2020; Kose & Arslan, 2016; Lin et al., 2013; Pai et al., 2021; Rodzman et al., 2020; Walkington & Bernacki, 2018; D. Xu & Wang, 2005), and decreased subject-related anxiety (Hwang et al., 2020).

Hwang et al. (2020) utilized a Personalized Virtual Learning Environment (PVLE) known as the Intelligent eLearning System (IeLS). This system was designed to enhance students’ mathematical learning by customizing the course content to their needs. Hwang and colleagues (2020) conducted a quasi-experimental study comparing student mathematics learning between the IeLS and a non-personalized Virtual Learning Environment (VLE). Their sample consisted of 228 fourth-grade students sampled from six classes in a Taiwanese school, randomly assigned to either the experimental (IeLS) or control group (VLE). The IeLS system implemented fuzzy logic to monitor and record student activities, such as browsing paths, time spent on tasks, quiz scores, and document interactions. The IeLS system further distinguished two experimental groups. The first group consisted of students whose IeLS measured their cognitive performance (assessment performance) and their affective states (i.e., concentration, willingness to learn, and patience with the learning material). In contrast, the other experimental group’s IeLS only measured their cognitive performance. In both groups, the IeLs then chose the most appropriate content based on the students’ performance and created personalized learning plans for each student. No significant differences were detected in the pre-test mathematic scores of the participants of the experimental and control groups.

Hwang et al. (2020) found that students using IeLS achieved significantly higher scores on quizzes and the final exam than those using the non-personalized IeLS, especially as the course progressed. Moreover, students in the IELS group that measured cognitive and affective components scored significantly higher on the final exam than those whose IELS only considered cognitive performance. Furthermore, students in the two experimental groups had significantly lower math-related anxiety than the VLE condition, but they observed no statistically significant differences in post-test, math-related anxiety in the two experimental groups. Their findings suggest that using AI to personalize math learning leads to better performance and lower math-related anxiety overall. Also, AI-empowered online learning systems should include affective states, such as concentration, patience, and willingness to learn, in addition to assessments to optimize their benefits.

Another study examined how AI-powered learning systems can influence university students’ (n=110) performance in three computer science courses (Kose & Arslan, 2016). In this experiment, Kose and Arslan (2016) divided the participants randomly into either an experimental or control group, with the experimental group using a LMS using artificial neural networks (ANN) trained by the Cognitive Development Optimization Algorithm (CoDOA). This specific study used adaptive assessment to tailor the content for each student, thereby creating a range of personalized learning paths for the courses. Students who used the intelligent e-learning system achieved significantly higher grades in three computer programming courses compared to a control group that received only traditional face-to-face lectures. In addition, students felt that the system improved their comprehension of the subject and heightened their enjoyment and engagement with the learning material. Moreover, the students thought that the LMS accurately provided them with relevant content to master the subjects. This study extends the benefits of AI, as shown in Hwang et al. (2020), to university students engaging with different subject matter, i.e., computer science courses. Moreover, it provides evidence of the effectiveness of an AI-powered LMS in a higher education, blended learning environment.

Another study by Walkington and Bernacki (2018) investigated the effects of personalized learning on student performance and engagement in algebra. The researchers designed an intervention within an ITS for algebra called Cognitive Tutor Algebra (CTA). The study involved 106 Algebra I high school students assigned to three experimental groups, with differing levels of dynamic assessment driving the students’ content personalization. The control group received standard algebra story problems on the LMS. The first experimental group received problems testing the same content as the control group. However, the CTA provided problems related to each student’s interests (sports, video games, or food). The algebra problems in this experimental group were customized to match the student’s interests, but they did not demonstrate how the algebra concepts could be used in those areas of interest. The experimental group in the second experiment received personalized algebra problems demonstrating how algebra concepts are applied in their area of interest (i.e., deep personalization). The researchers gathered information on student performance, including the proportion of correct first attempts at writing algebraic expressions and the number of correct expressions written per minute. They also looked at student engagement, such as their tendency to “game the system,” problem interestingness ratings, and attentional states like boredom and concentration. Additionally, they assessed the degree of quantitative engagement with the students’ interests.

Students in the personalized conditions (both surface and deep combined) performed better overall than those in the control condition on measures of correct first attempts (p = 0.076) and correct attempts per minute (p = 0.028). They were also less likely to game the system (p = 0.002) and rated problems more interesting (p = 0.029). Deep personalization was more effective for students with high engagement, while surface personalization was more effective for students with low engagement (interaction effect, p = 0.0499). Deep personalization was also associated with lower boredom (p = 0.021), regardless of engagement level. Overall, the findings suggest that personalized learning paths increase engagement, interest, and the academic performance of high school students. In addition, it shows how, based on students’ engagement levels, course creators can use different levels of personalized learning paths (deep or surface-level) to achieve better learning results.

Note, however, that creativity plays a role in producing better academic outcomes in learning, and an AI-power LMS holds the key to boosting students’ creativity. In a study by Lin et al., (2013), researchers examined the impact of a personalized creativity learning system (PCLS) on college students’ creativity learning. The system utilized a decision tree algorithm to tailor learning paths for students. The study involved 92 college students, aged 18 to 26 years, who participated in a computerized experiment in a laboratory setting. Participants completed three game-based creativity tasks (living room, kitchen, and bathroom scenarios), with each scenario consisting of 10 problems to “escape” the situation, earning one point per correct solution, with a maximum score of 30 points. The PCLS collected demographic data, self-perception of creativity, and learning styles and assigned participants to one of six learning paths. Data were analyzed using a C4.5 decision tree algorithm, along with gain ratio calculations, to identify relationships between student characteristics and creativity outcomes.

The AI adaptation in the PCLS was implemented through a multi-agent system comprising a user interface agent, creativity game agent, path agent, and questionnaire agent. The system personalizes the learning experience by adjusting the learning paths based on the student’s college major. For example, the decision tree algorithm recommends the “living room – kitchen – bathroom” path for science majors and the “bathroom – living room – kitchen” path for engineering majors. Students following the AI-suggested paths had a 90% chance of achieving an above-average creativity score, compared to a 52% chance for those on randomly assigned paths. The decision tree algorithm significantly predicted creativity scores, with students from the College of Science performing the best. Lin and colleagues (2013) concluded that AI-driven personalization could enhance creative learning but noted limitations, including the small sample size.

In summary, the reviewed studies collectively demonstrate the significant potential of AI-powered personalized learning systems in enhancing student motivation, engagement, and academic performance and reducing subject-related anxiety. Whether through adaptive learning environments like the IeLS in Hwang et al. (2020), which not only improved math performance but also reduced anxiety, or systems like the one in Kose and Arslan (2016), which tailored content to computer science students using artificial neural networks, AI adaptation led to measurable improvements in learning outcomes. Moreover, Walkington and Bernacki (2018) highlighted how personalized learning based on student interests improved algebra performance and engagement, with deep personalization further reducing boredom. Finally, Lin et al. (2013) emphasized the creative learning benefits provided by a decision tree algorithm-based system, which personalized learning paths according to student majors, resulting in enhanced creativity scores. Across these diverse contexts, AI’s role in personalizing education improved academic results and engaged students in a way that traditional methods often failed to achieve. These findings underscore the transformative potential of AI in education while also pointing to the need for further research to address limitations such as small sample sizes and to explore the generalizability of these systems to other educational settings and disciplines. We now turn our focus to how AI can help identify at-risk students.

3.1.2.1. Individualized Learning Paths: Identifying At-Risk Students and Providing Support and Feedback

Traditional classrooms, especially those characterized by high student-to-teacher ratios, pose significant challenges for at-risk students. Given the global learning crisis, it is now entirely possible that students with different levels of content mastery are grouped in the same course, making it difficult for course administrators to cater to each student’s needs. In online classes, students can achieve different levels of mastery. Course creators need to offer individualized learning paths to address the needs of students at risk of failing and dropping out.

Several sources (e.g., El-Bishouty et al., 2015; Libbrecht et al., 2015; Prinsloo et al., 2020), emphasize using predictive models in learning analytics to identify at-risk students preemptively. These models utilize historical data on student performance, engagement patterns, and other relevant factors to forecast the likelihood of a student struggling or failing a course. AI-powered systems can serve as early warning systems, particularly during a student’s first year, to detect those struggling academically or at risk of dropping out (see the review of Zawacki-Richter et al., 2019). These systems leverage academic performance, attendance, engagement with online learning platforms, and demographic information to identify potential warning signs (see the review of Zawacki-Richter et al., 2019). A personalized learning path can, based on early warning signs and identifying at-risk students, provide content, feedback, and assessments that support these learners.

In the section on dynamic assessments, I discuss the empirical evidence of how these support systems look in online or blended learning settings. I do this because identifying at-risk students includes harnessing the results produced by continually assessing individual students’ needs and levels.

3.1.2.2. Individualized Learning Paths and Self-Directed Learning

The idea of self-directed learning (SDL) lies in cultivating self-directed students. Self-directed students are those who possess the skills to plan, organize, and direct their learning (Kizilcec et al., 2017). Kizilcec et al. (2017) provide a stellar overview of how SDL looks in online learning, specifically MOOCs. In this section, I briefly summarize Kizilcec et al.’s (2017) overview to provide a glimpse into the usefulness AI holds for boosting SDL by referring to the six SRL strategies measured in their study: (a) goal setting, (b) strategic planning, (c) self-evaluation, (d) task strategies, (e) elaboration, and (f) help-seeking.

One aspect of SDL crucial in online learning is goal setting (Kizilcec et al., 2017). Goal settings include why students decide to follow an online course and breaking their goals down into smaller subgoals that would break down the behaviors necessary to achieve these goals. The idea of personalized learning is that learners have control over their learning paths, and course creators can facilitate personalized learning by clarifying learner’s goals beforehand. AI creates student-centered learning environments emphasizing students’ choice and control over course goals given their interests, future career paths, and needs (c.f., Bhutoria, 2022; Majeed, 2023; Zhang & Aslan, 2021). Some higher education systems implement a variation of this by allowing LMS to access students’ performance in previous courses and predict their goals and performance in a current yet-to-be-started course.

Another aspect of SDL involves how the course content is sequenced and structured in an online course (Kizilcec et al., 2017). Here, the choice of pace and content is not only the purview of learner profiles created by the data captured by LMSs (see this section) but also the learner’s preferences (Bhutoria, 2022; Zhang & Aslan, 2021). The idea of the feedback and support lies in the capabilities of AI in providing learners with real-time analytics about their strengths and weaknesses to help them gauge their progress toward their end goal (Klašnja‐Milićević et al., 2017; Majeed, 2023; Tang et al., 2021). Characteristically, a self-directed student will self-evaluate throughout a course to see if they are on track to achieve their learning goals (Kizilcec et al., 2021). However, with the added capabilities of AI, this characteristic of self-directed learners is available to students who were not previously considered self-directed learners. Furthermore, this allows the latter students to achieve a level of mastery over the course content that is usually associated with more successful self-directed learners.

Another component of SDL is that self-directed students develop strategies to regulate how they spend their study time (Effeney et al., 2013) and persist in these tasks despite any difficulties they encounter in the learning process (Richardson et al., 2012). Kizilcec et al. (2017) refer to this component of SRL in online courses as task strategies. By adapting content to their current level of mastery, AI-powered LMSs allow students of online courses to do this automatically. However, by providing personalized feedback and specific suggestions for improvement, an AI-powered LMS allows students to engage in SDL by choosing their tasks strategically to reach their learning goals successfully. This capability of AI will enable students who are not characteristically viewed as self-directed learners to engage in optimal task strategies that they may not have done before the incorporation of AI into online learning platforms. I discuss the implications of this with reference to the findings of Kizilcec et al.’s (2017) study, below.

Kizilcec et al. (2017) considered the last two strategies: elaboration and help-seeking. Elaboration refers to deepening one’s understanding of the course material by linking it with prior knowledge and trying to reflect on the implications of the learned material (Kizilcec et al., 2017). Help-seeking refers to all the behaviors students use to better understand the material they struggle with, whether through reaching out to the instructor or other students, or consulting extra or external learning materials (Kizilcec et al., 2017). The importance of help-seeking and elaboration are delineated more effectively in the section of this report dealing with how AI can provide continuous support and real-time feedback to learners (see here).

3.1.2.2.1. Empirical Evidence

In the studies I include below, I try to give a brief overview of the findings of Kizilcec et al. (2017). I only consider this study, given the dearth of literature studying SDL in online learning settings (Wong et al., 2019). For more information, I refer the reader to other meta-analyses or systematic reviews investigating SDL in online courses (see: Jansen et al., 2019; Wong et al., 2019; Xu, Zhao, Liew, et al., 2023; Xu, Zhao, Zhang, et al., 2023). Echoing the results of Kizilcec et al. (2017) that I discuss below, the majority of studies show how the use of SDL strategies is considered crucial for students’ achievement success in online and blended learning settings(Z. Xu, Zhao, Liew, et al., 2023).

Kizilcec et al. (2017) studied how various self-regulated learning (SRL) strategies looked at the learning behavior of 4,831 online learners enrolled in six distinct MOOCs offered in Spanish by Pontificia Universidad Católica de Chile through Coursera. Their study also identified individual traits that could predict weaker SRL skills to inform the design of learning environments that offer personalized support. The online courses encompassed a variety of subjects (e.g., education, management, engineering, and transportation). As a result, the sample comprised a diverse group of participants with varying ages, genders, education levels, employment statuses, and prior experience with online learning. Kizilcec et al. (2017) gathered survey data, including self-directed learning strategies, individual characteristics, and platform activity logs, which tracked interactions with course content and overall course achievement throughout the course duration. The performance activity logs provided additional information about the learning process by collecting data on students’ lecture viewing, assessment attempts, and review of previously studied course content. The researchers measured six SDL strategies (i.e., goal setting, strategic planning, self-evaluation, task strategies, elaboration, and help-seeking) and included a survey that collected data on student demographics, time commitment, prior experience with the course topic and online courses, course intentions, and enrollment motivations.

The first goal of this study was to ascertain how SDL strategies influenced the achievement of learning goals (i.e., watching lectures, completing assessments, and obtaining a course certificate). These three learning goals were used as outcomes assessed in this study’s analyses. The study’s second goal was to examine how SDL strategies presented themselves in the behaviors of the course attendants.  To explore this, Kizilcec et al. (2017) investigated how learners interacted with course content, identifying shifts between different interaction states (e.g., from completing a lecture to attempting an assessment). They also assessed the connections between SRL strategies and transition probabilities as well as per-session activity metrics, including time spent on activities, number of materials interacted with, and time between sessions. Kizilcec et al.’s (2017) third goal was to model students’ individual characteristics that predict their use of specific SRL strategies. Their penalized regression models considered 27 learner characteristics, including demographics, experience, commitment, goals, and motivations.

Kizilcec et al. found that goal-setting and strategic planning positively predicted students’ attainment of all three personal course goals. Conversely, help-seeking was negatively associated with goal attainment. Regarding how SRL strategies influence students’ engagement with online learning content, the researchers found that students with stronger SRL skills (except for help-seeking) frequently revisited previously completed materials, particularly assessments. Moreover, those inclined to seek help were less likely to pass an evaluation after completing a lecture. Lastly, this study by Kizilcec et al. revealed individual differences that influence SRL in an online learning setting. Older learners, those with higher levels of education (particularly Ph.D. holders), and those who had completed more online courses reported higher levels of most SRL strategies. More specifically, goal setting emerged as an important SRL strategy when differentiating between students who have started but not finished many online courses and students who have successfully finished many online courses. Students who had previously completed multiple online courses were actively involved in setting goals both before and during the courses they attended in this study. Likewise, employed students were more inclined to use several SRL strategies, including goal setting, strategic planning, and help-seeking. In contrast, school or university students consistently reported lower use of SRL strategies.

The study also shed light on how the motivations behind taking the course influenced the use of SRL strategies. Motivations for enrolling in a course that indicated a relevant and supportive life context—such as the course being beneficial for work, school, or research—were linked to more substantial SRL skills. In contrast, motivations that suggested a less supportive context, such as taking the course for enjoyment, facing a challenge, experiencing a MOOC, or changing careers, were associated with weaker SRL skills.

Overall, the study provides ample evidence on how course creators ought to consider how they allow AI-powered online courses to help their students set goals and strategically plan their individualized learning path. Moreover, what I, as well as Kizilcec and colleagues (2017), found interesting was that help-seeking was a negative predictor of course outcomes. The relation between help-seeking behaviors and unsuccessful online learning may be moderated by the extent to which online learning provides help to the learners (Kizilcec et al., 2017). This finding may signal a need for more supportive feedback mechanisms within online courses, whether through ITS or chatbots. In fact, this is, at least, one of the reasons why I chose to discuss help and feedback, which is a crucial component of AI’s promise in online learning settings. It is also possible that the students’ help-seeking behaviors in this study reflected problems in understanding the material. Besides, when they voiced their struggles, it may have been too late to salvage a successful course completion. From my perspective, this would mean an onus on course creators to provide early and ongoing interactive feedback to ensure the learning path becomes attainable for those struggling.

Additionally, the non-collaborative nature of MOOCs may also, according to Kizilcec et al. (2017), be a reason behind the relationship between help-seeking and unsuccessful course attainment in online courses. Therefore, this makes it imperative for course creators to consider ways to boost collaboration. One way, as I briefly mentioned above, is to include collaborative spaces in virtual worlds, whereas a more practical way could be to ensure chat forums and social media spaces where course students can meet, share ideas, and help each other.

Interestingly, students who were successful in the courses were those who engaged in frequent assessments of the topics they learned (Kizilcec et al., 2017). The findings of Koedinger et al. (2015) echo this by showing how students of online learning who engage frequently in online learning activities and assessments show better academic performance than students who passively experience the course content (e.g., watching videos). Moreover, it also provides empirical justification for my discussion of dynamic assessments.

Finally, the study of Kizilcec et al. (2017) suggested the importance of learner demographics in predicting the use of SRL strategies that underlie course success. This is not the only study to indicate this, preview literature reviews and studies also suggest the importance of student characteristics in predicting the successful completion of an online course (e.g., Martin et al., 2018; Munir et al., 2022; Pursel et al., 2016; Sønderlund et al., 2019). Other studies on retention in MOOC or online courses do not find that demographics are statistically significant predictors of outcomes(Hone & El Said, 2016). Nevertheless, it is important for course creators to collect data on demographics, and based on this, they use AI’s predictive capabilities to identify students who are less prone to succeed and provide frequent and interactive support to ensure that their courses yield high course completion rates. I stress this importance because the effect of individual differences on course outcomes may be particular to the course. Besides, as more students enroll in a specific course, its ML algorithms may be more attuned to identifying whether demographics influence course outcomes. As a result, course creators may be more responsive to the effect of individual differences if they exist in their online courses.

3.1.2.3. Individualized Learning Paths and The Increased Accessibility Associated With It

AI-powered online courses can provide quality education to learners in underserved areas. In the first report of CourseClout, I delineated the current nature of the global education crisis. Students from underdeveloped countries cannot access the same quality of teaching as students in developed countries. Moreover, in that report, I highlighted how teachers are sometimes inadequately trained to provide instruction in the courses they supervise. Given that online education is not bound by geographical area or time, it allows students from all over the world, including these underserved areas, to access quality instruction (Bhutoria, 2022; Inderjeet & Bhardwaj, 2024; H. Li & Dong, 2022; Majeed, 2023). Nevertheless, course creators who aim to serve these communities should also be aware of some challenges, such as the language of instruction not being in the student’s home language. Moreover, a lack of access to devices and the internet that would allow them to attend an online course (Maghsudi et al., 2021) remains a challenge that organizations, such as UNESCO, are actively working on fixing.

CourseClout uses generative AI technologies that can allow course creators to create AI instructors that deliver course content. This use of generative AI reduces the cost and time associated with producing online lecture videos. It can also increase the course content’s adaptability to the learners’ language. For example, AI tools, such as HeyGen, provide accurate translation services to up to 178 languages. This means that when course creators want to export their products to new countries, they can create a course instructor that delivers the course material in the student’s home language. Notably, mother tongue instruction remains an important component of academic achievement. For example, in a study by Bernhofer and Tonin (2022), students who took an exam in a language other than their mother tongue presented, on average, with 9.5% lower grades than those who completed the exam in their mother tongue. Of course, this study only relates to exam assessments, but it also implicates the importance of the content and ongoing assessments in online courses being in students’ mother tongues.

AI also provides increased accessibility for students who are blind, deaf, find traditional education difficult, and have learning disabilities, such as autism spectrum disorder (ASD; Almusaed et al., 2023; Dhananjaya et al., 2024; Kellems et al., 2020; Majeed, 2023; McMahon et al., 2016; Owoseni et al., 2024). AI-powered screen readers and voice assistants are already transforming how blind people access information and interact with technology. One way in which blind students can be accommodated with AI in online learning lies in the ability of AI to adapt the content to the individual learner. AI can adapt learning materials and assessments to accommodate the needs of blind students, such as providing alternative formats like Braille or audio recordings. For example, the DIAGRAM Center, supported by the U.S. Department of Education, now focuses on leveraging AI to make digital content accessible for students who are blind (U.S. Department of Education, 2017). In a similar way, the content can be adapted or augmented to help those who are deaf. For example, AI can be used to develop more robust and accurate speech recognition systems, enabling individuals who are blind to interact with computers and other devices using their voice (c.f., Belpaeme et al., 2018). Other examples include Presentation Translator, a PowerPoint add-in that generates real-time subtitles for presentations, benefiting students who are deaf or hard of hearing (Rangaiah, 2021). This technology enables students to follow along with lectures and presentations, even if they cannot hear the speaker.

It is also possible that AI helps students with communication disorders. While Dhananjaya et al. (2024) mention this possibility, they fail to address exactly how this may happen in practice. Nevertheless, it is easy to infer its usefulness. For example, speech recognition tools tailored to varied speech patterns enhance these students’ communication accuracy and allow them to correct their mistakes in real-time. At the same time, text-to-speech (TTS) and speech-to-text (STT) functionalities may help them to participate in audio and text discussions with other course attendees. Moreover, predictive text and autocomplete features support fluency, facilitating quicker, more effective communication, and given the individual nature of ML algorithms, the benefits of these features are becoming more accurate. Additionally, AI-driven augmentative and alternative communication (AAC) tools can enable students with severe speech impairments to express themselves through symbols and phrases. Another feature lies in the rise of ITS and the chatbots driving them. The chatbots can reduce the students’ reliance on verbal or complex written responses by offering guided, individualized support, and gesture recognition systems allow non-verbal interaction options. These AI advancements make online education more inclusive, allowing students with speech disorders to engage fully in their learning environments.

AI can also help students who struggle with certain subjects in traditional education studies. For example, a study by Pai et al. (2021) employed an ITS that allowed them to interact and understand mathematical concepts they struggled with. Overall, the learners’ performance improved significantly. Moreover, in some subjects, such as geography, students at the foundational level struggle to effectively use spatial reasoning to understand the geometrical relationships between objects (Beisenbayeva et al., 2024). Beisenbayeva et al. (2024) aptly note how struggles at the foundational levels influence these students’ performance at later levels in a geography subject. For example, students find it more challenging to construct mental models and to understand the relationship between the abstract concepts in the course. We discuss this study using the empirical evidence below.

3.1.2.3.1. Empirical Evidence 

Overall, the studies discussed below show how AI, specifically in conjunction with VR/AR technologies, can improve the learning of students who struggle with subject-specific concepts (Beisenbayeva et al., 2024) as well as students with learning disabilities (S. Dutt & Ahuja, 2024; Morris et al., 2022)s), intellectual disabilities (IDs: Safak & Yavuz, 2024), and ASD (McMahon et al., 2016).

Beisenbayeva et al. (2024) studied 82 tenth-grade students to ascertain the influence of AR on their spatial reasoning abilities and academic performance in a geography subject. The researchers randomly assigned subjects to either a control or experimental group. The former group received traditional face-to-face instructions on the subject, whereas the latter received a Geometrica app on their mobile devices for five 40-minute geometry lessons over two weeks. The Geometrica app provided a simulated world where the learners interacted with the objects to increase their understanding of the spatial relationships between them. Before the experiment, the groups had equivalent scores regarding their geometrical competencies. The researchers found that the experimental group showed a significantly improved understanding of the subject matter compared to the control group. Moreover, they were more engaged and enjoyed the learning process more than the control group.

In another study, Dutt and Ahuja (2024) designed an ITS that would take into account the needs of learners (n = 83) with learning disabilities (LD). Their ITS was also used to measure the students’ levels of engagement and emotions (e.g., frustration) during the learning process. Given this, these results also serve as empirical evidence for the section on learning engagement and enjoyment (see here). Overall, the students gave favorable ratings for the system. Moreover, the study found that using an ITS in an online course environment for students with LD reduced the cognitive load associated with traditional ways of learning and improved the course outcomes for these participants.

Morris et al. (2022) aimed to determine the effectiveness of a combined approach using point-of-view video modeling (POV-VM), explicit instruction (EI), and AR in teaching math skills to students with disabilities. The researchers sought to understand how this intervention affected student performance on targeted math skills, their ability to maintain those skills, and their capacity to transfer the learned skills to applied word problems. The study used a multiple-baseline design to assess student performance over time. The researchers found that the students’ performance immediately improved due to their intervention. Moreover, the positive effects of the intervention on the students’ math skills held even after the intervention stopped. The authors concluded that the effectiveness of this study lies in how it allowed these students to engage in SRL and take control of their individualized learning path. Also, by engaging multiple senses, the intervention allowed equitable participation for these students and allowed the researcher to provide content to these students based on their individual sensorial needs.

Another study by McMahon et al. (2016) assessed using AR to teach science vocabulary to college students with intellectual disabilities (IDs) and ASD. Three sets of science-related vocabulary words were targeted: human bones, organs, and cell biology, with ten words in each set. McMahon et al. (2016) developed vocabulary tests to assess students’ ability to define and label these terms. They created an AR intervention using the Aurasma app on iPads. This intervention involved printed markers (vocabulary cards) that triggered short videos containing definitions, images, and 3D simulations of the vocabulary words when scanned with the app. The results showed that all four students effectively learned the three science vocabulary terms through the AR intervention. Also, visual analysis indicated that the students showed a significant improvement in their ability to define and label scientific terms after using the AR intervention. The use of AI-power AR seemed to be useful for enhancing the performance of students with ID and ASD.

Mentioning more studies, in detail,  are beyond the scope of this report. In this paragraph, I briefly summarize the findings of other studies. Safak and Yavuz (2024) examined the effectiveness of Augmented Reality (AR) combined with the Constant Time Delay (CTD) method in teaching students with IDs about human organs and their functions. Results showed that all participants improved their ability to identify organs and describe functions, retained their knowledge over several weeks, and successfully applied their understanding to new contexts, with teachers affirming the method’s practicality and engagement benefits. In another study, Kellems et al. (2020) evaluated the effectiveness of an Augmented Reality (AR) intervention package in teaching procedural math skills to eighth-grade students with specific LDs, focusing on Common Core-based problem-solving. Results showed a functional relationship between the AR intervention and improved accuracy in completing problem-solving steps, with students achieving high accuracy levels post-intervention, though only partial skill retention over time. Social validity feedback indicated that students found the AR intervention helpful and enjoyable, suggesting that AR can enhance engagement and procedural learning, with broader implications for using AI and AR/VR in special education to personalize learning and support accessibility.

In conclusion, these studies underscore the potential of AI, VR, and AR technologies in enhancing accessibility and learning outcomes for students with disabilities, including those with learning disabilities, intellectual disabilities, and ASD. The results suggest that integrating interactive, immersive technologies can improve students’ understanding of complex concepts, reduce cognitive load, boost engagement, and enable personalized learning pathways that cater to individual needs. For course designers aiming to increase accessibility, these findings highlight the importance of incorporating adaptive AI and immersive tools tailored to diverse learning profiles. This approach can support equitable learning experiences and encourage students to engage more fully with content, ultimately leading to more effective and inclusive education environments.

3.1.2.4. Dynamic and Adaptive Assessments: The Catalyst For Feedback and Support.

Traditional assessments, even those implemented in online courses without dynamic assessment, assess all learners despite their individual differences in the same way. In this sense, these assessments do not allow questions or tasks on tests to differ based on the individual student’s understanding of the content, prior knowledge, strengths, or weaknesses. In contrast, the ability of AI to handle large amounts of data allows it to analyze an array of data about a student’s response times, error patterns, and interaction with the learning material (Costa et al., 2021; Liu & Yu, 2022). In turn, the ability to interpret this vast amount of data allows the AI to continually update its understanding of a learner’s individual needs and to adapt the content to the learner (Costa et al., 2021; Liu & Yu, 2022). Moreover, with the advent of ITS systems, students can interact with virtual assistants that provide them with continual feedback and clarification and allow students to correct their misunderstanding of course concepts and reinforce their understanding (Almusaed et al., 2023; Libbrecht et al., 2015; Owoseni et al., 2024).

AI achieves dynamic assessments through an array of mechanisms. ITS are systems that use AI algorithms to provide personalized instruction and feedback to learners. These systems can analyze student responses, identify misconceptions, and offer targeted hints, explanations, or prompts to guide their understanding (Belpaeme et al., 2018; Munir et al., 2022). Regarding assessments, these systems analyze student responses to identify performance in real-time and adapt these assessments and exercises accordingly (Costa et al., 2021; Murtaza et al., 2022). Assessment may, for example, become progressively more difficult for students who can answer correctly, whereas it may change the level of questions or provide hints to those students who are struggling.

This real-time adaptation of assessments is a practical scenario in online and blended-learning settings for the first time due to AI. By personalizing and sequencing the next assessment item or exercise, ITS ensures that students stay appropriately challenged throughout the learning process, thereby reducing frustration and boredom, and increasing engagement (Inderjeet & Bhardwaj, 2024; Luan et al., 2020; Maghsudi et al., 2021). Later in the report, I link this with the control-value theory (CVT) of achievement emotions in education. Incorporating AI into ITS also allows a learning path where students are provided immediate feedback on their answers, identify areas for improvement, and suggest additional learning resources to ensure deep learning (Almusaed et al., 2023). As chatbots are increasingly incorporated with ITS, learners can truly deepen their knowledge of subjects anywhere and anytime (Han et al., 2022). This, in turn, has implications for students’ subject-level anxiety, retention, and SDL, which we explore in the evidence below.

So far, it is clear that AI increases student support. Another way support is facilitated is through the automation of grading (Inderjeet & Bhardwaj, 2024).  For example, Lu et al. (2021)developed a system that automatically generates tests using natural language processing. Automated grading holds the promise for students to get to the root of their course-related weaknesses quicker (Inderjeet & Bhardwaj, 2024), and, with the added capabilities of ITSs, students can now more readily engage in SDL strategies to master the learning content.

The benefits of all this dynamic assessment, feedback, and support lie in how it truly personalized the learning experience (Majeed, 2023; Outman et al., 2023), increases student motivation and engagement (Akavova et al., 2023; Ouyang et al., 2023), allows for target remediation (Dhananjaya et al., 2024; Zawacki-Richter et al., 2019), as well as the efficient allocation of time and resources for both students, educators, and course creators (Dhananjaya et al., 2024; Prinsloo et al., 2020). In the following section, we explore the evidence.

3.1.2.4.1. Empirical Evidence 

A meta-analysis by Steenbergen-Hu and Cooper (2014) included 39 studies published between 1990 and 2011 that assessed the effect of ITS on students’ academic performance. The sample sizes for these studies ranged from 20 to 1,066 participants, and the included studies represented a diverse set of ITSs used. The meta-analysis examined 22 types of ITSs, including frequently studied ones like AutoTutor, ALEKS, xTex-Sys, and WISE, implemented in various instructional contexts: as primary instructors, as a part of a blended-learning context, as supplemental instruction, as support during lab-based activities, and as homework support. The control groups of the included studies included traditional classroom instruction, printed or digital reading materials, computer-assisted instruction, self-guided learning, control groups, and human tutoring.

Steenbergen-Hu and Cooper (2014) found that the ITS had a moderate, positive impact on college students’ academic learning. The average effect size ranged from g = 0.32 to g = 0.37, indicating that students using ITS achieved better academic performance than those in comparison groups. An interactive ITS that employs dynamic assessment, support, and feedback is more effective than students who were in controls that employed self-reliant learning, reading of the printed text, traditional face-to-face instruction, and non-AI powered, computer-assisted learning.

Another meta-analysis by Fletcher and Kulik (2016) sought to determine how effective Intelligent Tutoring Systems are at improving student performance. The researchers analyzed over 500 sources on ITSs and selected 50 that met their inclusion criteria. The ITS studies included real-time feedback, support, and content adjustment capabilities that utilize the promise of including AI-powered dynamic assessments throughout the learning process. The meta-analysis found that ITSs are more effective than conventional classes, other forms of computer tutoring, and human tutoring (Fletcher & Kulik, 2016). The students who used ITS outperformed students in traditional courses 92% of the time, and the learning improvements were statistically significant in 78% of the cases. The average effect size for ITS, measured as Hedges g, was 0.61. In contrast, other forms of computer tutoring had an average effect size of 0.31, and human tutoring had an average effect size of 0.40. The effect of ITS on improved academic performance signifies a large effect size, with Fletcher and Kulik (2016) concluding that ITS “…can provide unusually effective instruction.” (p.9) and “…may represent a breakthrough for digital tutoring technology.” (p. 9).

Throughout the sections of dynamic assessments above, I emphasized the importance of quick and immediate feedback as a vital support characteristic of AI-powered ITS. The empirical substantiation for immediate feedback comes from a study by Singh et al. (2011), which examined the effectiveness of a web-based ITS used by 68 eighth-grade students. Singh et al. (2011) implemented a repeated measures design, randomly assigning the student to either the Immediate Feedback with Tutoring (IFT) condition or the Business as Usual (BAU) condition. Students in both groups completed pre-tests and post-tests to assess learning gains. The IFT group used the ASSISTments platform for their homework, receiving immediate feedback, tutoring, and hints, while the BAU group completed paper-based homework and received feedback the following day. To assess the quality of teacher feedback, researchers videotaped the second day’s homework review sessions. Finally, researchers analyzed data from the pre-tests and post-tests to determine outcomes. Students in the IFT condition learned more from their homework than students in the BAU condition. This emphasizes the importance of immediate feedback. Furthermore, their findings suggest that online course creators can bolster the effects of immediate feedback by including tutoring within such use cases.

In another study, Han et al. (2022) investigated the use of AI chatbots as a support tool for nursing students (n = 61)  learning about electronic fetal monitoring (EFM). Specifically, the study aimed to establish how a chatbot affects various aspects of learning, including knowledge acquisition, SDL strategies,  retention, and critical reasoning skills. In this study, the control group only watched the video lectures, whereas the experimental group attended the video lectures and accessed the chatbots for further learning. Han et al. (2022) found that while the experimental group showed increased interest, engagement, and SDL, the groups did not show statistically different results regarding knowledge acquisition and critical reasoning skills. While this study suggests no effect of ITS on academic performance, it can only be viewed as a primary study. Instead, it remains helpful to consider that the meta-analyses above point to an overall benefit that this study, alone, cannot negate. However, this study also signifies how contextual differences might affect the effectiveness of ITS assessment, feedback, and support. This possibility is something that course creators ought to consider.

3.1.2.5. Immersive Learning Experiences: Interactivity, Engagement, and Motivation

Interactivity may come from how the interaction between the student and content becomes more engaging (e.g., through gamification or VR) due to using AI and how AI facilitates the interactions between students and their instructors. Our discussion of how AI augments content for an individual learner is already the first point of interaction between the student and the content (Murtaza et al., 2022; Owoseni et al., 2024) and the first time where the student is allowed to feel unique enough that the content is adapted based on their profile of needs, strengths, goals, and prior knowledge.

Another way in which AI boosts student interaction with content is by providing an immersive experience through gamification (Hooshyar et al., 2015; Inderjeet & Bhardwaj, 2024; U.S. Department of Education, 2017) and simulations via VR/AR technologies (Annuš, 2024; Bhutoria, 2022; Dhananjaya et al., 2024; Owoseni et al., 2024). By incorporating AI into an online LMS, course creators can allow the AI to create educational games that challenge students to apply their knowledge and skills in engaging and rewarding ways (Inderjeet & Bhardwaj, 2024). Inderjeet and Bhardwaj (2024) suggest that by including gamification, the learning process is boosted by making the process more engaging and fun for learners. However, Inderjeet and Bhardwaj (2024) caution against the over-gamification of the learning process as they suggest that this could negate deeper learning and understanding of the course material. It is, therefore, important for course creators to consider a balance between gamification and activities that would allow for a deeper integration of the course material.

By incorporating AI into LMS, course creators can provide learners with simulations to enhance their learning and retention of the course material. This enhancement rests on AI’s ability to generate realistic scenarios, based on the course content, where students can apply their knowledge (Inderjeet & Bhardwaj, 2024). Like those offered by Labster (2024), AI-powered simulations can provide realistic, interactive experiences that mimic real-world scenarios, such as laboratory experiments or medical procedures. These simulations allow learners to practice skills, make decisions, and experience the consequences of their actions in a risk-free environment (Inderjeet & Bhardwaj, 2024). These simulations can occur without the use of VR/AR technologies. Still, in most of the reviewed literature, educative simulations are usually created and administered via these technologies.

VR/AR technologies offer immersive, interactive learning experiences. VR technologies can allow students to take virtual field trips to places like ancient Rome or explore complex subjects such as the human body in 3D (c.f., Bhutoria, 2022). On the other hand, AR overlays digital information in the real world, providing real-time feedback and guidance. These technologies make abstract concepts tangible, particularly in fields like mathematics and anatomy, where spatial reasoning is crucial (Dhananjaya et al., 2024; Inderjeet & Bhardwaj, 2024). AI enhances these VR and AR experiences by personalizing content and adjusting difficulty levels, allowing students to engage deeply with the material. For instance, VR simulations have proven effective in training social skills for children with autism, providing a safe, controlled environment to practice real-life interactions (see Lorenzo et al., 2016).

Creating immersive and engaging learning environments is urgently needed, especially in light of the global education crisis. Across different cultures, the role of boredom in predicting dire academic outcomes holds consistently (Frenzel et al., 2007; Luszczynska et al., 2005; Pekrun & Goetz, 2023). What is troubling is that international research has shown boredom to be the most commonly experienced emotion by students in education(Goetz et al., 2006; Gu et al., 2023; Martz et al., 2018). A meta-analysis by Tze et al. (2016) found that boredom had the most substantial negative relationship with academic motivation (r = -.40), followed by study strategies and behaviors (r = -.35), and lastly, achievement (r = -.16). Moreover, a report by the Bill and Melinda Gates Foundation has linked the high high-school dropout rates in America to the boredom experienced by these students in the classroom(Bridgeland et al., 2006). For a more in-depth understanding of this emotion in the context of the global learning crisis, please refer to CourseClout’s first research report.

However, CVT (Pekrun, 2006; Pekrun et al., 2002) provides a theoretical framework for understanding and addressing boredom by increasing students’ engagement in online courses. Boredom occurs during learning when students perceive a task as having low personal value (Pekrun & Goetz, 2024). Here, the personalized learning paths and AI’s ability to match the content with the goals and needs of the student, provide a way to circumvent boredom in online courses. Moreover, the interactive nature of VR/AR technologies also provides a way in which previously impersonal, abstract material can become personally relevant for these students. Despite increased personal value, boredom may also occur when students perceive the material as either providing no or too much of a challenge to them (Pekrun & Goetz, 2024). In this way, adaptive assessments allow for the challenge to adjust to the student’s level of mastery over the content. So, in a way, this report already showed how course creators can alleviate boredom by using AI. Moreover, in the paragraphs above, I also made the case for how the immersive nature of VR/AR technologies could further add to the interactivity provided by adaptive AI.

Interactivity in education, however, goes beyond the interaction between learners and the course content to include their interactions with educators and peers. Targeted support for individual needs enhances the interaction between course teachers and students. This streamlines communication between students and teachers and enables teachers to provide support in the best way possible. Moreover, with the inclusion of chatbots, students now have 24/7 interactive support that alleviates the pressure on the student-teacher relationship, allowing teachers more time to focus on high-level actions that produce results in their course. AI further bolsters the collaboration between students by, for example, grouping students with similar skills, weaknesses, or learning strategies together and making the collaborative nature of education more productive (c.f., Jia & Yang, 2022).

3.1.2.5.1. Empirical Evidence

Let us first consider the empirical evidence from the studies already mentioned. The study by Walkington and Bernacki (2019) investigated using an AI-powered ITS that personalized math problems to students’ interests. The students found the personalized problems more engaging. Students who received problems demonstrating how mathematics applies to their interests showed significantly less boredom than those in the control group. Most studies (e.g., Kose & Arslan, 2016; Ouyang et al., 2023; Walkington & Bernacki, 2019; D. Xu & Wang, 2005) we discussed where personalization increased academic performance show that the AI-powered LMS provided more engaging and enjoyable experiences. In light of the discussion on CVT, the mechanisms of engagement and lack of boredom to produce better academic outcomes become clear.

CourseClout is a company that provides a unique spin to online learning. They aim to produce quality, high-production-level video content that uses immersive graphics, such as Pixar animations, to produce a higher level of learner engagement and knowledge retention. The company’s focus is on narrative-driven content, which allows students to become immersed in the course’s story. The final goal is to, in their words, produce bingeable learning. It is, therefore, inevitable to look at the evidence suggesting this approach’s usefulness.

Fan et al. (2015) acknowledged the boring nature of science classes and implemented story-based scenarios into a virtual library environment. They wanted the course content to be more engaging, fun, and relatable. An example of what they did includes a scenario where students role-play as an adventurer trying to cross a valley using a vine and actively seek solutions for the problems presented in the story. The researchers embedded personalization by allowing the students to choose different stories based on their interests.

The study examined 31 tenth-grade students who learned basic dynamics-related concepts (Fan et al., 2015). The researchers employed the Technology Acceptance Model (TAM) and the Test of Integrated Science Process Skills (TIPS) to evaluate student acceptance and skill development. The TAM questionnaire assessed students’ perceived usefulness, ease of use, and intention to use the system. First, students completed the TIPS, received instruction on using the story-based virtual experiment environment, and then engaged with the system for 50 minutes, solving problems across four story scenes. Finally, they completed the TAM questionnaire.

The results show that students considered the story-based virtual environment helpful for learning Physics concepts, supporting the idea that the system is effective for education. Also, the researchers found a significant positive correlation between perceived usefulness and students’ intention to use the system, implying that a system perceived as valuable will likely encourage continued use. No significant correlation was found between student grades (past Physics grades and TIPS scores) and their acceptance of the system. This finding suggests the system appeals to students across different academic achievement levels. This, therefore, provides a preliminary justification for CourseClout’s use of story-based, narrative learning to appeal to different types of learners in online courses.

Karamert and Kuyumcu Vardar (2021)incorporated gamification in the experimental group’s mathematics lessons. The authors utilized a quasi-experimental design with a pretest-posttest control group on a sample of 46 5th-grade students, matching the experimental and control groups based on their academic averages in mathematics. The researchers introduced gamification in different ways. For example, each student in the experimental group received a progress map, which they personalized. This map visually represented the students’ progress through the “fractions” unit. In addition, the researchers allowed students to choose and design the avatars they use in the progress maps. Finally, students earned badges based on their weekly achievements on the course content. The control group received teaching as usual.

While the study did not utilize AI for gamification, its results still shed light on its usefulness. Moreover, with the added benefit of personalization, AI can increase the effects of gamification. The achievement test results found a statistically significant difference, favoring the experimental group. This indicates that gamification positively impacts students’ academic performance in mathematics. While both the experimental and control groups experienced an increase in their achievement test scores, the increase in the experimental group was significantly larger.

While we already acknowledge the usefulness of VR/AR technologies, it might be useful to delineate the findings of a comprehensive systematic review of Fernández-Batanero et al. (2022) that investigated the impact of AR on students with educational needs. They identified and reviewed the results of 18 studies published between 2016 and 2021. The samples included students from every level of education, from primary school up until higher education and the student samples consisted of students with IDs, ASD, LDs, and hearing impairment. Their findings on the benefits of AR’s immersive capabilities for students with special needs are unmistakable. Overall, AR has produced better academic results, engagement, motivation, communication, social interaction, and autonomy for students, similar to students in traditional learning settings.

Enhancing communication was one thing we discussed as an important way in which AI can facilitate the social aspects of courses. To this end, Yang et al. (2013) highlighted the important way in which social environments explain the high dropout rates in MOOCs. Yang and colleagues (2013) employed a survival model on the social data (e.g., discussions in online forums) to understand the duration of time until students drop out of MOOCs. They, specifically, looked at students’ posting behaviors (i.e., starting a post, length of interactions, and duration of posting) and social network behaviors (i.e., students’ social positioning in discussions, betweenness, and closeness). The sample for the study consisted of 771 users who had posted at least once on the discussion forums of the literature course by the 7th week of the course. Students who engaged earlier and longer in discussion forums were more likely to complete an MOOC successfully. Moreover, students with higher authority scores, a metric indicating their influence and engagement in discussions, were also less likely to drop out. This finding emphasizes the importance of students taking an active role in facilitating discussions and sharing their knowledge with others, which in turn strengthens their connection to the course and the learning community. It is, therefore, imperative for course creators to consider ways to stimulate collaboration and social interaction in online courses.

3.2. Enhanced Educator or Creator Capabilities 

AI is reshaping the capabilities available to educators and course creators in the rapidly evolving digital education landscape. This section delves into two key areas where AI proves invaluable: video creation and enhanced educator flexibility. AI tools enable creators to produce high-quality, accessible video content at significantly reduced costs, thereby expanding course reach and engagement. Additionally, automating routine tasks, like grading and plagiarism detection, allows educators to devote more time to meaningful student interactions and creative lesson planning. By integrating AI-driven systems, platforms like CourseClout offer transformative resources for educators, making their roles more efficient and dynamic. This report highlights the powerful applications of AI in these domains, showcasing how CourseClout can support educators in leveraging these technologies to maximize course impact.

3.2.1. Video

It is important to note that AI now allows course creators to create videos at a fraction of a cost and at high production value (Liu & Yu, 2023). We, also, referred to the beneficiality of AI video creation by referring to its capability to increase the accessibility of courses by real-time language transaltion. CourseClout aims to facilitate this process by producing high-quality and binge-able course videos. Notably, however, we leave this section short as delineating how CourseClout achieves this borders on proprietary information. We encourage the reader to reach out to CourseClout to understand how they can help leverage the power of AI to enhance video content.

3.2.2. Enhanced Flexibility 

Teachers are bombarded with a plethora of activities that stretch their availability to attend to matters that may help individuals with their course struggles. Grading, plagiarism detection, attendance registers, and personalized feedback are all things that can now be automated with AI(Almusaed et al., 2023; Inderjeet & Bhardwaj, 2024; Klašnja‐Milićević et al., 2017; Outman et al., 2023; Yildirim & Celepcikay, 2021). If course creators and educators automate these tasks, they allow for the benefits of adaptive learning paths and dynamic assessments. Moreover, by freeing up their time, teachers now have enhanced flexibility that allows them to interact with students, provide remedial support, and plan ways to increase the utility of the course’s content (Baker & Smith, 2019; Zohuri, 2024).

AI-powered LMSs have the added benefit of allowing teachers to plan, create, and augment lesson content(Akavova et al., 2023; Almusaed et al., 2023; Owoseni et al., 2024). It may help teachers create lesson plans based on available resources and provide additional resources to students when they are most needed (Akavova et al., 2023; Owoseni et al., 2024). A seminal book by Owoseni et al. (2024) explores how teachers and course creators can use tools such as ChatGPT to facilitate their tasks.

Lesson planning and content generation has become a walk in the park, thanks to tools such as ChatGPT. Teachers can use ChatGPT to quickly generate detailed lesson outlines, including learning objectives, materials, procedures, and assessments (Owoseni et al., 2024). This can save teachers a significant amount of time and effort, allowing them to focus on the more creative and interactive aspects of lesson planning. Moreover, teachers can now explore ways to use ChatGPT to generate engaging and diverse course materials, such as slideshows, quizzes, case studies, and interactive exercises (Owoseni et al., 2024). For a more detailed delineation, please refer to the book of Owoseni et al. (2024).

In this report, I have included ways in which AI can make the assessment process more useful and how AI’s different capabilities can boost teachers’ use of this platform. While research on AI’s usefulness focuses mostly on the learner, its usefulness for course creators is unmistakable. Moreover, CourseClout’s experience makes them a valuable partner in this endeavour.

The integration of AI in education extends beyond mere convenience; it empowers educators to amplify their teaching methods and content creation. With the support of AI-driven video production and automated administrative tasks, educators can concentrate on enhancing student engagement and developing innovative learning experiences. CourseClout’s approach exemplifies how AI can be used to augment these capabilities, providing educators with flexible, impactful tools tailored to modern educational demands. This report illustrates the advantages AI brings to the education sector, emphasizing the partnership potential with CourseClout for those seeking to enhance their course delivery and learner engagement through AI.

Chapter 4: Conclusion

The report illustrates how AI, through its integration in Education 4.0, is transforming both student learning and educator capabilities in online education. With unprecedented advances in AI, machine learning, and big data, educators and course creators can deliver tailored, immersive, and adaptive learning experiences that support individual needs, drive engagement, and help students overcome learning challenges. Through personalized learning paths, real-time feedback, and advanced tools like VR and AR, AI is not only making learning accessible and engaging but also fostering the autonomy and growth of students.

CourseClout’s AI-driven approach to course creation and management exemplifies this transformation. By enabling video creation at scale, facilitating adaptive assessments, and automating administrative tasks, CourseClout allows educators to focus on high-value interactions with learners and improve the quality of their courses. As the demand for online learning continues to grow, partnerships with platforms like CourseClout provide an invaluable opportunity for course creators to maximize the impact of their educational content and contribute meaningfully to the evolving landscape of digital education. Through AI, the goal of equitable, effective, and learner-centered education is becoming an achievable reality.

References

Akavova, A., Temirkhanova, Z., & Lorsanova, Z. (2023). Adaptive learning and artificial intelligence in the educational space. E3S Web of Conferences, 451, 06011. https://doi.org/10.1051/e3sconf/202345106011

Al-Badia, A., Khan, A., & Eid-Alotaibi. (2022). Perceptions of learners and instructors towards artificial intelligence in personalized learning. Procedia Computer Science, 201, 445–451. https://doi.org/10.1016/j.procs.2022.03.058

Albert, A., & Hallowel, M. R. (2013). Revamping occupational safety and health training: Integrating andragogical principles for the adult learner. Construction Economics and Building, 13(3), 128–140. https://doi.org/10.5130/AJCEB.v13i3.3178

Almusaed, A., Almssad, A., Yitmen, I., & Homod, R. Z. (2023). Enhancing student engagement: Harnessing “AIED”’s power in hybrid education—A review analysis. Education Sciences, 13(7), 632. https://doi.org/10.3390/educsci13070632

Annuš, N. (2024). Education in the Age of Artificial Intelligence. TEM Journal, 404–413. https://doi.org/10.18421/TEM131-42

Awad, S. O. I., Mohamed, Y., & Shaheen, R. (2022). Applications of artificial intelligence in education. 1(1). https://doi.org/xxx 0000/j cicom.vol1no1.9

Baker, R. S. J. D. (2007). Modeling and understanding students’ off-task behavior in intelligent tutoring systems. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, 1059–1068. https://dl.acm.org/doi/10.1145/1240624.1240785

Baker, T., & Smith, L. (2019). Educ-AI-tion Rebooted? Exploring the future of artificial intelligence in schools and colleges (p. 55). Nesta.

Barsalou, L. W. (2008). Grounding symbolic operations in the brain’s modal systems. In G. R. Semin & E. R. Smith (Eds.), Embodied grounding: Social, cognitive, and neuroscientific approaches. Cambridge University Press.

Beisenbayeva, G. K., Mubarakov, A. M., Seylova, Z. T., Zhadrayeva, L. U., & Artymbayeva, B. N. (2024). Evaluating the Impact of an Augmented Reality App on Geometry Learning in Kazakh Secondary Schools. Journal of Information Technology Education: Research, 23, 022. https://doi.org/10.28945/5355

Belpaeme, T., Kennedy, J., Ramachandran, A., Scassellati, B., & Tanaka, F. (2018). Social robots for education: A review. Science Robotics, 3(21), eaat5954. https://doi.org/10.1126/scirobotics.aat5954

Bernard, J., Chang, T.-W., Popescu, E., & Graf, S. (2016). Improving learning style identification by considering different weights of behavior patterns using particle swarm optmization. In Y. Li, M. Chang, M. Kravcik, E. Popescu, R. Huang, Kinshuk, & N.-S. Chen (Eds.), State-of-the-art and future directions of smart learning (pp. 39–50). Springer.

Bernhofer, J., & Tonin, M. (2022). The effect of the language of instruction on academic performance. Labour Economics, 78, 102218. https://doi.org/10.1016/j.labeco.2022.102218

Bezhovski, Z., & Poorani, S. (2016). The Evolution of E-Learning and New Trends.

Bhutoria, A. (2022). Personalized education and Artificial Intelligence in the United States, China, and India: A systematic review using a Human-In-The-Loop model. Computers and Education: Artificial Intelligence, 3, 100068. https://doi.org/10.1016/j.caeai.2022.100068

Bridgeland, J. M., DiIulio, J. J., & Morison, K. B. (2006). The silent epidemic: Perspectives of high school dropouts [Research  report]. Civic Enterprises in association with Peter D. Hart Research Associates for the Bill & Melinda Gates Foundation.

https://docs.gatesfoundation.org/Documents/TheSilentEpidemic3-06FINAL.pdf

Camacho-Morles, J., Slemp, G. R., Pekrun, R., Loderer, K., Hou, H., & Oades, L. G. (2021). Activity Achievement Emotions and Academic Performance: A Meta-analysis. Educational Psychology Review, 33(3), 1051–1095. https://doi.org/10.1007/s10648-020-09585-3

Carbonell, J. R. (1970). AI in CAI: An Artificial-Intelligence Approach to Computer-Assisted Instruction. IEEE Transactions on Man-Machine Systems, 11(4), 190–202. https://doi.org/10.1109/TMMS.1970.299942

Cascales-Martínez, A., Martínez-Segura, M.-J., Pérez-López, D., & Contero, M. (2016). Using an Augmented Reality Enhanced Tabletop System to Promote Learning of Mathematics: A Case Study with Students with Special Educational Needs. EURASIA Journal of Mathematics, Science and Technology Education, 13(2). https://doi.org/10.12973/eurasia.2017.00621a

Cathy Li & Farah Lalani. (2020, April 29). The COVID-19 pandemic has changed education forever. This is how. World Economic Forum: Health and Health Care Systems. https://www.weforum.org/agenda/2020/04/coronavirus-education-global-covid19-online-digital-learning/

Clay, J., & Orwig, C. J. (1999). Your learning style and language learning (Version 3.5). Summer Institute of Linguistics, Inc. (SIL) International.

Cooper, S., & Sahami, M. (2013). Reflections on Stanford’s MOOCs. Commun. ACM, 56(2), 28–30. https://doi.org/10.1145/2408776.2408787

Costa, R. S., Tan, Q., Pivot, F., Zhang, X., & Wang, H. (2021). Personalized and adaptive learning: Educational practice and technological impact. Linguagem e Tecnologia, 14(3), e33445. https://doi.org/10.35699/1983-3652.2021.33445

Coursera Staff. (2024, September 18). Augmented Reality vs. Virtual Reality: What’s the Difference? Coursera. https://www.coursera.org/articles/augmented-reality-vs-virtual-reality

Crompton, H., & Burke, D. (2023). Artificial intelligence in higher education: The state of the field.

International Journal of Educational Technology in Higher Education, 20(1), 22. https://doi.org/10.1186/s41239-023-00392-8

Dhananjaya, G. M., Goudar, R. H., Rathod, V. N., & Hukkeri, G. S. (2024). A Digital Recommendation System for Personalized Learning to Enhance Online Education: A Review. IEEE Access, 12, 34019–34041. https://doi.org/10.1109/ACCESS.2024.3369901

D’Mello, S., Craig, S., Witherspoon, A., McDaniel, B., & Graesser, A. (2008). Automatic detection of learner’s affect from conversational cues. User Model. User-Adapt. Interact., 18, 45–80. https://doi.org/10.1007/s11257-007-9037-6

Dutt, A., Ismail, M. A., & Herawan, T. (2017). A Systematic Review on Educational Data Mining. IEEE Access, 5, 15991–16005. https://doi.org/10.1109/ACCESS.2017.2654247

Dutt, S., & Ahuja, N. J. (2024). “Intelligent tutoring effects on induced emotions and cognitive load of learning-disabled learners.” Disability and Rehabilitation: Assistive Technology, 1–15. https://doi.org/10.1080/17483107.2024.2357685

Dwivedi, Y. K., Kshetri, N., Hughes, L., Slade, E. L., Jeyaraj, A., Kar, A. K., Baabdullah, A. M., Koohang, A., Raghavan, V., Ahuja, M., Albanna, H., Albashrawi, M. A., Al-Busaidi, A. S., Balakrishnan, J., Barlette, Y., Basu, S., Bose, I., Brooks, L., Buhalis, D., … Wright, R. (2023). Opinion Paper: “So what if ChatGPT wrote it?” Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI for research, practice and policy. International Journal of Information Management, 71, 102642. https://doi.org/10.1016/j.ijinfomgt.2023.102642

Effeney, G., Carroll, A., & Bahr, N. (2013). Self-Regulated Learning: Key strategies and their sources in a sample of adolescent males. Australian Journal of Educational and Developmental Psychology, 13, 58–74.

El-Bishouty, M. M., Chang, T., Lima, R., Thaha, M. B., Kinshuk, & Graf, S. (2015). Analyzing learner characteristics and courses based on cognitive abilities, learning styles, and context. In M. Chang & Y. Li (Eds.), Smart learning environments (pp. 3–25). Springer.

Fan, M. X., Kuo, R., Chang, M., & Heh, J. S. (2015). Story-based virtual experiment environment. In Smart learning environments (pp. 175–198). Springer.

Felder, R., & Silverman, L. (1988). Learning and teaching styles. Journal of Engineering Education, 94(1), 674–681.

Ferguson, R. (2012). Learning analytics: Drivers, developments and challenges. International Journal of Technology Enhanced Learning, 4(5/6), 304. https://doi.org/10.1504/IJTEL.2012.051816

Fernández-Batanero, J. M., Montenegro-Rueda, M., & Fernández-Cerero, J. (2022). Use of Augmented Reality for Students with Educational Needs: A Systematic Review (2016–2021). Societies, 12(2), 36. https://doi.org/10.3390/soc12020036

Fletcher, J. D., & Kulik, J. A. (2016). Effectiveness of Intelligent Tutoring Systems A Meta-Analytic Review. Review of Educational Research, 86(1), 42–78. https://doi.org/10.3102/0034654315581420

Frenzel, A., Thrash, T., Pekrun, R., & Goetz, T. (2007). Achievement Emotions in Germany and China A Cross-Cultural Validation of the Academic Emotions Questionnaire—Mathematics. Journal of Cross-Cultural Psychology, 38, 302–309. https://doi.org/10.1177/0022022107300276

Giannakos, M., Sharma, K., Pappas, I., Kostakos, V., & Velloso, E. (2019). Multimodal data as a means to understand the learning experience. International Journal of Information Management. https://doi.org/10.1016/j.ijinfomgt.2019.02.003

Gobert, J. D., & Sao Pedro, M. A. (2017). Digital assessment environments for scientific inquiry practices. In A. A. Rupp & J. P. Leighton (Eds.), The Wiley Handbook of Cognition and Assessment (pp. 508–534). Wiley. doi: 10.1002/9781118956588

Goetz, T., Frenzel, A. C., Pekrun, R., & Hall, N. C. (2006). The Domain Specificity of Academic Emotional Experiences. The Journal of Experimental Education, 75(1), 5–29.

https://doi.org/10.3200/JEXE.75.1.5-29

Gómez-Pulido, J. A., Park, Y., Soto, R., & Lanza-Gutiérrez, J. M. (2023). Data Analytics and Machine Learning in Education. Applied Sciences, 13(3), 1418. https://doi.org/10.3390/app13031418

Gu, Z., Yang, C., Su, Q., & Liang, Y. (2023). The boredom proneness levels of Chinese college students increased over time: A meta-analysis of birth cohort differences from 2009 to 2020. Personality and Individual Differences, 215, 112370. https://doi.org/10.1016/j.paid.2023.112370

Han, J.-W., Park, J., & Lee, H. (2022). Analysis of the effect of an artificial intelligence chatbot educational program on non-face-to-face classes: A quasi-experimental study. BMC Medical Education, 22. https://doi.org/10.1186/s12909-022-03898-3

Hinojo-Lucena, F.-J., Aznar-Díaz, I., Cáceres-Reche, M.-P., & Romero-Rodríguez, J.-M. (2019). Artificial Intelligence in Higher Education: A Bibliometric Study on its Impact in the Scientific Literature. Education Sciences, 9(1), 51. https://doi.org/10.3390/educsci9010051

Hone, K. S., & El Said, G. R. (2016). Exploring the factors affecting MOOC retention: A survey study. Computers & Education, 98, 157–168. https://doi.org/10.1016/j.compedu.2016.03.016

Hooshyar, D., Ahmad, R., Yousefi, M., Fathi, M., Horng, S.-J., & Lim, H. (2015). Applying an Online Game-based Formative Assessment in a Flowchart-based Intelligent Tutoring System for Improving Problem-Solving Skills. Computers & Education, 94, 18–36. https://doi.org/10.1016/j.compedu.2015.10.013

Hwang, G.-J., Sung, H.-Y., Chang, S.-C., & Huang, X.-C. (2020). A fuzzy expert system-based adaptive learning approach to improving students’ learning performances by considering affective and cognitive factors. Computers and Education: Artificial Intelligence, 1, 100003. https://doi.org/10.1016/j.caeai.2020.100003

Inderjeet, & Bhardwaj, R. (2024). AI Integrated Model Design for Education System for Analytical Computation (Mathematics). 2024 4th International Conference on Advance Computing and Innovative Technologies in Engineering (ICACITE), 1412–1417.

https://doi.org/10.1109/ICACITE60783.2024.10616503

Jansen, R. S., van Leeuwen, A., Janssen, J., Jak, S., & Kester, L. (2019). Self-regulated learning partially mediates the effect of self-regulated learning interventions on achievement in higher education: A meta-analysis. Educational Research Review, 28, 100292. https://doi.org/10.1016/j.edurev.2019.100292

Jia, Y., & Yang, L. (2022). Online Interactive Teaching System of Sanda Course Based on P2P Network. In W. Fu & G. Sun (Eds.), E-Learning, e-Education, and Online Training (pp. 103–117). Springer Nature Switzerland.

Jordan, K. (2015). MOOC completion rates: The data [Researcher’s personal website]. Katy Jordan: Researching Education and Technology. http://www.katyjordan.com/MOOCproject.html#:~:text=To%20switch%20between%20charts%20showing,much%20more%20typical%20MOOC%20size.

Karamert, Ö., & Kuyumcu Vardar, A. (2021). The effect of gamification on young mathematics learners’ achievements and attitudes. Journal of Educational Technology and Online Learning, 4(2), 96–114. https://doi.org/10.31681/jetol.904704

Kellems, R. O., Eichelberger, C., Cacciatore, G., Jensen, M., Frazier, B., Simons, K., & Zaru, M. (2020). Using Video-Based Instruction via Augmented Reality to Teach Mathematics to Middle School Students With Learning Disabilities. Journal of Learning Disabilities, 53(4), 277–291. https://doi.org/10.1177/0022219420906452

Kizilcec, R. F., Makridis, C. A., & Sadowski, K. C. (2021). Pandemic response policies’ democratizing effects on online learning. Proceedings of the National Academy of Sciences, 118(11), e2026725118. https://doi.org/10.1073/pnas.2026725118

Kizilcec, R. F., Pérez-Sanagustín, M., & Maldonado, J. J. (2017). Self-regulated learning strategies predict learner behavior and goal attainment in Massive Open Online Courses. Computers in Education, 104(1), 18–33. https://doi.org/10.1016/j.compedu.2016.10.001

Klašnja‐Milićević, A., Ivanović, M., & Budimac, Z. (2017). Data science in education: Big data and learning analytics. Computer Applications in Engineering Education, 25(6), 1066–1078. https://doi.org/10.1002/cae.21844

Knowles, M. S., Holton, E. F., & Swanson, R. A. (2011). The adult learner: The definitive classic in adult education and human resource development. Elsevier.

Koedinger, K. R., Kim, J., Jia, J. Z., McLaughlin, E. A., & Bier, N. L. (2015). Learning is Not a Spectator Sport: Doing is Better than Watching for Learning from a MOOC. Proceedings of the Second (2015) ACM Conference on Learning @ Scale, 111–120. https://doi.org/10.1145/2724660.2724681

Kose, U., & Arslan, A. (2016). Intelligent E-Learning System for Improving Students’ Academic Achievements in Computer Programming Courses. International Journal of Engineering Education, 32(1(A)), 185–198.

Krenn, M., Pollice, R., Guo, S. Y., Aldeghi, M., Cervera-Lierta, A., Friederich, P., dos Passos Gomes, G., Häse, F., Jinich, A., Nigam, A., Yao, Z., & Aspuru-Guzik, A. (2022). On scientific understanding with artificial intelligence. Nature Reviews Physics, 4(12), 761–769. https://doi.org/10.1038/s42254-022-00518-3

Kucirkova, N., & Leaton Gray, S. (2023). Beyond Personalization: Embracing Democratic Learning Within Artificially Intelligent Systems. Educational Theory, 73(4), 469–489. https://doi.org/10.1111/edth.12590

Kuhail, M. A., Alturki, N., Alramlawi, S., & Alhejori, K. (2022). Interacting with educational chatbots: A systematic review. Education and Information Technologies, 28, 1–46. https://doi.org/10.1007/s10639-022-11177-3

Kumar, A., & Ahuja, N. J. (2020). An Adaptive Framework of Learner Model Using Learner Characteristics for Intelligent Tutoring Systems. In S. Choudhury, R. Mishra, R. G. Mishra, & A. Kumar (Eds.), Intelligent Communication, Control and Devices (pp. 425–433). Springer Singapore.

767.Sie, R. L., Delahunty, J., Bell, K., Percy, A., Rienties, B., Cao, T., & De Laat, M. (2018). Artificial Intelligence to Enhance Learning Design in UOW Online, a Unified Approach to Fully Online Learning. 2018 IEEE International Conference on Teaching, Assessment, and Learning for Engineering (TALE), 761–767. https://doi.org/10.1109/TALE.2018.8615283

Labster. (2024). 3D Learning Simulations. Labster. https://www.labster.com/3d

Le Grange, J. J. P., & Van Den Noortgate, W. (2024). Multilevel network meta-analysis of single-case experimental designs: A simulation study. [Master’s Thesis, KU Leuven. Faculteit Psychologie en Pedagogische Wetenschappen]. https://repository.teneo.libis.be:443/delivery/DeliveryManagerServlet?dps_pid=IE18506723

Li, C., Belter, M., Liu, J., & Lukosch, H. (2023). Immersive Virtual Reality Enabled Interventions for Autism Spectrum Disorder: A Systematic Review and Meta-Analysis. Electronics, 12(11), 2497. https://doi.org/10.3390/electronics12112497

Li, H., & Dong, X. (2022). Construction of Online Ideological and Political Education Platform Based on Artificial Intelligence Technology. In W. Fu & G. Sun (Eds.), E-Learning, e-Education, and Online Training (pp. 129–144). Springer Nature Switzerland.

Libbrecht, P., Müller, W., & Rebholz, S. (2015). Smart learner support through semi-automatic feedback. In Smart learning environments. Springer.

Lin, C. F., Yeh, Y., Hung, Y. H., & Chang, R. I. (2013). Data mining for providing a personalized learning path in creativity: An application of decision trees. Computers & Education, 68, 199–210. https://doi.org/10.1016/j.compedu.2013.05.009

Liu, C., & Yu, H. (2023). AI-Empowered Persuasive Video Generation: A Survey. ACM Comput. Surv., 55(13s), 285:1-285:31. https://doi.org/10.1145/3588764

Liu, M., & Yu, D. (2022). Towards intelligent E-learning systems. 28(7), 7845–7876. https://doi.org/10.1007/s10639-022-11479-6

Liu, X., & Ardakani, S. P. (2022). A machine learning enabled affective E-learning system model. Education and Information Technologies, 27(7), 9913–9934. https://doi.org/10.1007/s10639-022-11010-x

Lorenzo, G., Lledó, A., Pomares, J., & Roig, R. (2016). Design and application of an immersive virtual reality system to enhance emotional skills for children with autism spectrum disorders. Computers & Education, 98, 192–205. https://doi.org/10.1016/j.compedu.2016.03.018

Lu, O., Huang, A., Tsai, D., & Yang, S. (2021). Expert Authored and Machine Generated Short Answer Questions for Assessing Students Learning Performance. Educational Technology and Society, 24, 159–173.

Luan, H., Geczy, P., Lai, H., Gobert, J., Yang, S. J. H., Ogata, H., Baltes, J., Guerra, R., Li, P., & Tsai, C.-C. (2020). Challenges and future directions of big data and artificial intelligence in education. Frontiers in Psychology, 11, 580820–580820. https://doi.org/10.3389/fpsyg.2020.580820

Luckin, R., & Holmes, W. (2016). Intelligence Unleashed: An argument for AI in Education.

Lukianets, H., & Lukianets, T. (2023). PROMISES AND PERILS OF AI USE ON THE TERTIARY EDUCATIONAL LEVEL. Grail of Science, 25, 306–311. https://doi.org/10.36074/grail-of-science.17.03.2023.053

Luszczynska, A., Gutiérrez‐Doña, B., & Schwarzer, R. (2005). General self‐efficacy in various domains of human functioning: Evidence from five countries. International Journal of Psychology, 40(2), 80–89. https://doi.org/10.1080/00207590444000041

Lynch, T., & Ghergulescu, I. (2017). Large Scale Evaluation of Learning Flow (p. 64). https://doi.org/10.1109/ICALT.2017.98

Maghsudi, S., Lan, A., & Xu, J. (2021). Personalized Education in the AI Era: What to Expect Next? IEEE Signal Processing Magazine, 38(3), 37–50. https://doi.org/10.1109/MSP.2021.3055032

Majeed, M. (2023). Predicting the Future of Education in the Light of Artificial Intelligence. In C. Krishnan, M. Babbar, & G. Singh (Eds.), Digital Transformation in Education: Emerging Markets and Opportunities (pp. 173–188). BENTHAM SCIENCE PUBLISHERS.

https://doi.org/10.2174/9789815124750123010014

Marougkas, A., Troussas, C., Krouska, A., & Sgouropoulou, C. (2023). Virtual Reality in Education: A Review of Learning Theories, Approaches and Methodologies for the Last Decade. Electronics, 12(13), 2832. https://doi.org/10.3390/electronics12132832

Martin, F., Wang, C., & Sadaf, A. (2018). Student perception of helpfulness of facilitation strategies that enhance instructor presence, connectedness, engagement and learning in online courses. The Internet and Higher Education, 37, 52–65. https://doi.org/10.1016/j.iheduc.2018.01.003

Martindale, T., & Dowdy, M. (2010). Personal learning environments. Emerging Technologies in Distance Education, 7, 177–193.

Martín-Gutiérrez, J., Mora, C. E., Añorbe-Díaz, B., & González-Marrero, A. (2017). Virtual Technologies Trends in Education. EURASIA Journal of Mathematics, Science and Technology Education, 13(2). https://doi.org/10.12973/eurasia.2017.00626a

Martz, M. E., Schulenberg, J. E., Patrick, M. E., & Kloska, D. D. (2018). “I Am So Bored!”: Prevalence Rates and Sociodemographic and Contextual Correlates of High Boredom Among American Adolescents. Youth & Society, 50(5), 688–710. https://doi.org/10.1177/0044118X15626624

McMahon, D. D., Cihak, D. F., Wright, R. E., & Bell, S. M. (2016). Augmented Reality for Teaching Science Vocabulary to Postsecondary Education Students With Intellectual Disabilities and Autism. Journal of Research on Technology in Education, 48(1), 38–56. https://doi.org/10.1080/15391523.2015.1103149

Mikropoulos, T. A., & Natsis, A. (2011). Educational virtual environments: A ten-year review of empirical research (1999–2009). Computers & Education, 56(3), 769–780. https://doi.org/10.1016/j.compedu.2010.10.020

Morris, J. R., Hughes, E. M., Stocker, J. D., & Davis, E. S. (2022). Using Video Modeling, Explicit Instruction, and Augmented Reality to Teach Mathematics to Students With Disabilities. Learning Disability Quarterly, 45(4), 306–319. https://doi.org/10.1177/07319487211040470

Munir, H., Vogel, B., & Jacobsson, A. (2022). Artificial Intelligence and Machine Learning Approaches in Digital Education: A Systematic Revision. Information, 13(4), 203. https://doi.org/10.3390/info13040203

Murtaza, M., Ahmed, Y., Shamsi, J. A., Sherwani, F., & Usman, M. (2022). AI-Based Personalized E-Learning Systems: Issues, Challenges, and Solutions. IEEE Access, 10, 81323–81342. https://doi.org/10.1109/ACCESS.2022.3193938

Orlosky, J., Kiyokawa, K., & Takemura, H. (2017). Virtual and Augmented Reality on the 5G Highway. Journal of Information Processing, 25, 133–141. https://doi.org/10.2197/ipsjjip.25.133

Outman, H., Smail, A., Souad, A., Souad, A., & Hicham, A. (2023). Developments in Artificial Intelligence: What’s at stake for education? In M. Khaldi, M. Kaddouri, & M. Erradi (Eds.), Proceedings of the International Conference on E-Learning and Smart Engineering Systems (ELSES 2023) (Vol. 14, pp. 267–274). Atlantis Press International BV. https://doi.org/10.2991/978-94-6463-360-3_27

Ouyang, F., & Jiao, P. (2021). Artificial intelligence in education: The three paradigms. Computers and Education: Artificial Intelligence, 2, 100020. https://doi.org/10.1016/j.caeai.2021.100020

Ouyang, F., Wu, M., Zheng, L., Zhang, L., & Jiao, P. (2023). Integration of artificial intelligence performance prediction and learning analytics to improve student learning in online engineering course. International Journal of Educational Technology in Higher Education, 20(1), 4. https://doi.org/10.1186/s41239-022-00372-4

Owoseni, A., Kolade, O., & Egbetokun, A. (2024). Generative AI in Higher Education: Innovation Strategies for Teaching and Learning. Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-60179-8

Pai, K.-C., Kuo, B.-C., Liao, C.-H., & Liu, Y.-M. (2021). An application of Chinese dialogue-based intelligent tutoring system in remedial instruction for mathematics learning. Educational Psychology, 41(2), 137–152. https://doi.org/10.1080/01443410.2020.1731427

Palanisamy, P., Thilarajah, S., & Chen, Z. (2021). Providing equitable education through personalised adaptive learning. ASCILITE 2021: Back to the Future – ASCILITE ‘21 Proceedings ASCILITE 2021 in Armidale, 202–207. https://doi.org/10.14742/ascilite2021.0129

Papanastasiou, G., Drigas, A., Skianis, C., Lytras, M., & Papanastasiou, E. (2019). Virtual and augmented reality effects on K-12, higher and tertiary education students’ twenty-first century skills. Virtual Reality, 23(4), 425–436. https://doi.org/10.1007/s10055-018-0363-2

Pardamean, B., Suparyanto, T., Cenggoro, T. W., Sudigyo, D., & Anugrahana, A. (2022). AI-Based Learning Style Prediction in Online Learning for Primary Education. IEEE Access, 10, 35725–35735. https://doi.org/10.1109/ACCESS.2022.3160177

Pekrun, R. (2006). The control-value theory of achievement emotions: Assumptions, corollaries, and implications for educational research and practice. Educational Psychology Review, 18, 315–341.

Pekrun, R., & Goetz, T. (2023). How universal are academic emotions? A control-value theory persepective. In G. Hagenauer, R. Lazarides, & H. Järvenoja (Eds.), Motivation and Emotion in Learning and Teaching across Educational Contexts: Theoretical and Methodological Perspectives and Empirical Insights (1st ed., pp. 85–99). Routledge.

Pekrun, R., & Goetz, T. (2024). Boredom: A control-value theory approach. In M. Bieleke, W. Wolff, & C. S. Martarelli (Eds.), The Routledge International Handbook of Boredom (1st ed., pp. 75–89). Routledge. https://doi.org/10.4324/9781003271536

Pekrun, R., Goetz, T., Titz, W., & Perry, R. P. (2002). Academic Emotions in Students’ Self-Regulated Learning and Achievement: A Program of Qualitative and Quantitative Research. Educational Psychologist, 37(2), 91–105. https://doi.org/10.1207/S15326985EP3702_4

Pekrun, R., Lichtenfeld, S., Marsh, H. W., Murayama, K., & Goetz, T. (2017). Achievement Emotions and Academic Performance: Longitudinal Models of Reciprocal Effects. Child Development, 88(5), 1653–1670. https://doi.org/10.1111/cdev.12704

Popenici, S. A. D., & Kerr, S. (2017). Exploring the impact of artificial intelligence on teaching and learning in higher education. Research and Practice in Technology Enhanced Learning, 12(1), 22. https://doi.org/10.1186/s41039-017-0062-8

Precedence Research. (2024). AI in education market size, share, and trends 2024 to 2034 [Company website]. Precedence Research. https://www.precedenceresearch.com/ai-in-education-market

Prinsloo, P., Blaschke, L. M., & Olcott Jr., D. (2020). How Do We Know They are Learning? Student Data and the Synergies of Human and Artificial Intelligence (AI). EDEN Conference Proceedings, 1, 262–276. https://doi.org/10.38069/edenconf-2020-ac0025

Pursel, B. K., Zhang, L., Jablokow, K. W., Choi, G. W., & Velegol, D. (2016). Understanding MOOC students: Motivations and behaviours indicative of MOOC completion. Journal of Computer Assisted Learning, 32(3), 202–217. https://doi.org/10.1111/jcal.12131

Rahimi, R. A., & Oh, G. S. (2024). Rethinking the role of educators in the 21st century: Navigating globalization, technology, and pandemics. Journal of Marketing Analytics, 12(2), 182–197. https://doi.org/10.1057/s41270-024-00303-4

Rangaiah, M. (2021). Applications of AI in the education sector. Analytics Steps. https://www.analyticssteps.com/blogs

Richardson, M., Abraham, C., & Bond, R. (2012). Psychological Correlates of University Students’ Academic Performance: A Systematic Review and Meta-Analysis. Psychological Bulletin, 138, 353–387. https://doi.org/10.1037/a0026838

Rodzman, S. B. B., Abu Bakar, N., Choo, Y.-H., Ahmad Aljunid, S., Kamal Ismail, N., Abd Rahman, N., & Mohd Rosli, M. (2020). I-OnAR: A rule-based machine learning approach for intelligent assessment in an online learning environment. Indonesian Journal of Electrical Engineering and Computer Science, 17(2), 1021. https://doi.org/10.11591/ijeecs.v17.i2.pp1021-1028

ui, Z., & Badarch, T. (2022). Research on Applications of Artificial Intelligence in Education. American Journal of Computer Science and Technology, 5(2), 72. https://doi.org/10.11648/j.ajcst.20220502.17

S. B. Vinay. (2023). APPLICATION OF ARTIFICIAL INTELLIGENCE (AI) IN SCHOOL TEACHING AND LEARNING PROCESS- REVIEW AND ANALYSIS. https://doi.org/10.17605/OSF.IO/AERNV

Şafak, P., & Yavuz, M. (2024). The Effectiveness of Augmented Reality Offered with Constant Time Delay Procedure in Teaching the Functions of Organs to Students with Intellectual Disabilities. Ankara Üniversitesi Eğitim Bilimleri Fakültesi Özel Eğitim Dergisi, 1–20. https://doi.org/10.21565/ozelegitimdergisi.1391301

Sahlberg, P. (2009). Creativity and innovation through lifelong learning. Lifelong Learning in Europe, 14, 53–60.

Sanabria, J. C., & Arámburo-Lizárraga, J. (2017). Enhancing 21st century skills with AR: Using the gradual immersion method to develop collaborative creativity. EURASIA Journal of Mathematics, Science and Technology Education, 13(2), 487–501.

Schmidt, H. G., Van der Molen, H., Te Winkel, W., & Wijnen, W. H. F. W. (2009). Constructivist, problem-based learning does work: A meta-analysis of curricular comparisons involving a single medical school. Educational Psychologist, 44(4), 227–249. https://doi.org/10.1080/00461520903213592

Shah, D. (2020, February 5). Capturing the Hype: Year of the MOOC Timeline Explained — Class Central. The Report by Class Central. https://www.classcentral.com/report/mooc-hype-year-1/

Singh, R., Saleem, M., Pradhan, P., Heffernan, C., Heffernan, N. T., Razzaq, L., Dailey, M. D., O’Connor, C., & Mulcahy, C. (2011). Feedback during Web-Based Homework: The Role of Hints. In G. Biswas, S.

Bull, J. Kay, & A. Mitrovic (Eds.), Artificial Intelligence in Education (Vol. 6738, pp. 328–336). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-21869-9_43

SNS Insider Pvt. Ltd. (2023). Global next-generation ICT market size, share & segmentation by blockchain, artificial intelligence (AI), by metaverse, by cybersecurity, by software’s, by IOT, by region, and global forecast 2043-2032 (ICT Report No. SNS/ICT/4041; Information & Communication Technology, p. 800). https://www.snsinsider.com/reports/global-next-generation-ict-market-4041

Soflano, M., Connolly, T. M., & Hainey, T. (2015). An application of adaptive games-based learning based on learning style to teach SQL. Computers & Education, 86, 192–211. https://doi.org/10.1016/j.compedu.2015.03.015

Sønderlund, A. L., Hughes, E., & Smith, J. (2019). The efficacy of learning analytics interventions in higher education: A systematic review. British Journal of Educational Technology, 50(5), 2594–2618. https://doi.org/10.1111/bjet.12720

Spring, B., & Hitchcock, K. (2009). Evidence-Based Practice. In I. B. Weiner & W. E. Craighead (Eds.), The Corsini Encyclopedia of Psychology. John Wiley & Sons.

Steenbergen-Hu, S., & Cooper, H. (2014). A meta-analysis of the effectiveness of intelligent tutoring systems on college students’ academic learning. Journal of Educational Psychology, 106(2), 331–347. https://doi.org/10.1037/a0034752

Steglitz, J., Warnick, J. L., Hoffman, S. A., Johnston, W., & Spring, B. (2015). Evidence-Based Practice. In International Encyclopedia of the Social & Behavioral Sciences (pp. 332–338). Elsevier. https://doi.org/10.1016/B978-0-08-097086-8.10540-9

Stryker, C., & Kavlakoglu, E. (2024, August 9). What Is Artificial Intelligence (AI)? | IBM. https://www.ibm.com/topics/artificial-intelligence

Tang, K.-Y., Chang, C.-Y., & Hwang, G.-J. (2021). Trends in artificial intelligence-supported e- learning: A systematic review and co-citation network analysis (1998-2019). Interactive Learning Environments. https://doi.org/10.1080/10494820.2021.1875001

Tempelaar, D. T., Rienties, B., & Giesbers, B. (2015). In search for the most informative data for feedback generation: Learning analytics in a data-rich context. Computers in Human Behavior, 47, 157–167. https://doi.org/10.1016/j.chb.2014.05.038

Tze, V. M. C., Daniels, L. M., & Klassen, R. M. (2016). Evaluating the Relationship Between Boredom and Academic Outcomes: A Meta-Analysis. Educational Psychology Review, 28(1), 119–144. https://doi.org/10.1007/s10648-015-9301-y

U.S. Department of Education. (2017). Reimagining the role of technology in education: 2017 national education technology plan update (p. 106) [Technology Plan Update]. U.S. Department of Education.

Walkington, C., & Bernacki, M. (2018). Personalizing Algebra to Students’ Individual Interests in an Intelligent Tutoring System: Moderators of Impact. International Journal of Artificial Intelligence in Education, 29. https://doi.org/10.1007/s40593-018-0168-1

Walkington, C., & Bernacki, M. L. (2019). Personalizing Algebra to Students’ Individual Interests in an Intelligent Tutoring System: Moderators of Impact. International Journal of Artificial Intelligence in Education, 29(1), 58–88. https://doi.org/10.1007/s40593-018-0168-1

Wartman, S., & Combs, C. (2017). Medical Education Must Move From the Information Age to the Age of Artificial Intelligence. Academic Medicine, 93, 1. https://doi.org/10.1097/ACM.0000000000002044

Wong, J., Baars, M., Davis, D., Van Der Zee, T., Houben, G.-J., & Paas, F. (2019). Supporting Self-Regulated Learning in Online Learning Environments and MOOCs: A Systematic Review. International Journal of Human–Computer Interaction, 35(4–5), 356–373. https://doi.org/10.1080/10447318.2018.1543084

Xu, D., & Wang, H. (2005). Intelligent agent supported personalization for virtual learning environments. Decision Support Systems, 42(2), 825–843. https://doi.org/10.1016/j.dss.2005.05.033

Xu, Z., Zhao, Y., Liew, J., Zhou, X., & Kogut, A. (2023). Synthesizing research evidence on self-regulated learning and academic achievement in online and blended learning environments: A scoping review. Educational Research Review, 39, 100510. https://doi.org/10.1016/j.edurev.2023.100510

Xu, Z., Zhao, Y., Zhang, B., Liew, J., & Kogut, A. (2023). A meta-analysis of the efficacy of self-regulated learning interventions on academic achievement in online and blended environments in K-12 and higher education. Behaviour & Information Technology, 42(16), 2911–2931. https://doi.org/10.1080/0144929X.2022.2151935

Yang, D., Sinha, T., Adamson, D., & Rosé, C. P. (2013, November). ‘Turn on, Tune in, Drop out’: Anticipating student dropouts in MOOCs. Workshop on Data Driven Education. NIPS Workshop on Data Driven Education.

Yang, M., & Weng, F. (2023). AI-Powered Personalized Learning Journeys: Revolutionizing Information Management for College Students in Online Platforms. Journal of Information Systems Engineering and Management, 8(1), 23196. https://doi.org/10.55267/iadt.07.14079

Yildirim, Y., & Celepcikay, A. (2021). Artificial Intelligence and Machine Learning Applications in Education. Eurasian Journal of Higher Education, 2(4), 1–11. https://doi.org/10.31039/ejohe.2021.4.49

Yin, W. (2022). Personalized Hybrid Education Framework Based on Neuroevolution Methodologies. Computational Intelligence and Neuroscience, 2022, 1–16. https://doi.org/10.1155/2022/6925668

Yousafzai, A., Chang, V., Gani, A., & Noor, R. M. (2016). Multimedia augmented m-learning: Issues, trends and open challenges. International Journal of Information Management, 36(5), 784–792. https://doi.org/10.1016/j.ijinfomgt.2016.05.010

Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education – where are the educators? International Journal of Educational Technology in Higher Education, 16(1), 39. https://doi.org/10.1186/s41239-019-0171-0

Zhang, K., & Aslan, A. B. (2021). AI technologies for education: Recent research & future directions. Computers and Education: Artificial Intelligence, 2, 100025. https://doi.org/10.1016/j.caeai.2021.100025

Zhu, C. (2022). Intelligent Interactive Mobile Teaching Platform in Colleges and Universities Based on Artificial Intelligence Network. In W. Fu & G. Sun (Eds.), E-Learning, e-Education, and Online Training (pp. 145–157). Springer Nature Switzerland.

Zohuri, B. (2024). Revolutionizing Education: The Dynamic Synergy of Personalized Learning and Artificial Intelligence. International Journal of Advanced Engineering and Management Research, 09(01), 143–153. https://doi.org/10.51505/ijaemr.2024.9111