7. Big Data and Learning Analytics: Empowering Adaptive Classroom Learning

 

By Larry G. Martin  

Innovations in how educational and training organizations identify, collect, store, analyze, and report students’ digital data pave the way for Web3.0 personalized adaptive learning opportunities. Many higher education institutions are concerned that over 40.4 million Americans have enrolled in college but did not receive a degree (Bryant, 2024). Struggling academically in required courses, often in huge lecture halls, about 24% of first-time undergraduate students drop out of college, and 29.2% drop out after six years (Bryant, 2024). For students, traditional lecture-based classroom teaching is like reading a printed book with fixed content presented in the same order, regardless of the reader’s comprehension level. Classroom teachers, therefore, are seldom aware of the difficulties many students experience in these inflexible learning environments. Also, many organizations seek to make trainer-led employee training more productive and efficient in corporate training. A possible solution is found in the large volumes of big data generated as adults employ tools, e.g., computers, smartphones, Internet of Things devices, etc., from their digital learning hubs to access the digital apps from adaptive learning technologies sponsored by educational organizations (Peng & Spector, 2019).

Digital Apps Collect Diverse Sources of Student Data

Education and training organizations have adopted a variety of digital platforms that assist them in collecting and analyzing large volumes of data about learner interactions. When adults enroll in courses, they’re typically required to adopt several digital apps to assist with organizational enrollment (e.g., admissions, tuition and fees, and course registration) and educational experiences (e.g., course participation, levels of engagement, and academic performance). These apps generate vast amounts of diverse student data from various activities. For example, the Web3.0 Learning Management System (LMS) platform app can collect data from learners' interactions with course materials, assessments, and online discussions. Data from organization-based apps for Student Information Systems (SIS) include demographic, enrollment, academic performance, and financial records information. By monitoring students’ comments, likes, and shares related to educational content in Social Media apps and Online Forums, these organizations can gain insights into learners' preferences, interests, and peer interactions. Apps on students’ Wearable Devices can capture data on students’ physical activities, and data on students’ preferences for temperature, lighting, and other learning environmental conditions can be monitored via an organization’s intelligent Classroom Sensors. Lastly, Web3.0 Adaptive Learning Platforms routinely collect data on learners' interactions, preferences, and performance, which can be shared with educational providers. However, the data from these sources must be systematically collected to aid students’ learning processes.

Three Levels of Big Data Collected by Educational Organizations

Students’ digital footprints in the form of Big Data are collected and used at three broad levels of education and training organizations. At the microlevel, clickstream data are collected automatically during interactions between learners and their respective learning environments (e.g., coursework on LMSs and engagements with adaptive learning systems) (Fischer et al., 2020). Millions of second-by-second digital data points for each student provide detailed information on students’ cognitive strategies, affective states, or self-regulated learning (SRL) behaviors. Instructors can obtain actionable knowledge by assessing which actions are appropriate for different subgroups or student profiles (Fischer et al., 2020). At the Mesolevel, text data are systematically collected within minutes to hours during online writing activities. Obtained from online discussion forums, website databases, and social media interactions, these data provide insight into students’ levels of engagement, understanding, cognitive and social abilities, and affective states (Fischer et al., 2020). Educational organizations routinely collect and analyze Macrolevel organization-wide student data once or twice per term (Fischer et al., 2020). These data provide insight into each student’s demographic profile, admissions information, course enrollments, and degree completion. Although they have powerful potential, these data are powerless if not properly analyzed.

Educational Data Mining and Learning Analytics
Advancements in Web3.0-based learning analytics technologies have made it easier and cost-effective for educational organizations to use the enormous volumes of students’ data to benefit learners. Learning analytics involves measuring, collecting, analyzing, and reporting students' data to optimize learning experiences (Lee et al., 2020). The analysis of students’ data is like studying unorganized puzzle pieces. Educational data mining (EDM) provides a process for systematically organizing these unorganized data pieces into a complete, real-time, and accurate picture of learners and their learning efforts in specific classes. Using algorithms, EDM can discover hidden patterns in adult learners' behavior, learning styles, and preferences. They extract useful information from large datasets to identify patterns, relationships, and trends to create profiles of students' learning behaviors and patterns. These profiles are used to personalize and optimize the educational experiences of each learner.

Johnson et al. (2016) identified six key data points that are often included in the creation of learners’ profiles: demographic information (age, gender, ethnicity, and socio-economic background); academic records (grades, test scores, and completed courses); learning preferences, such as preferred learning styles (visual, auditory, and kinesthetic) and favored types of resources (videos, readings, and interactive simulations); behavioral data (time spent on tasks, participation in discussions, and engagement levels); psychological motivational traits (levels of self-efficacy); and social interactions (levels of collaboration and interaction with peers and instructors). The real-time profiles of each learner created from these data pave the way for innovative adaptive learning systems to customize courses with individual student characteristics (Zlatkovic et al., 2020).

Smart Adaptive Learning Systems

Intelligent adaptive learning systems are digital innovations designed to individualize and personalize the learning content so that students can experience self-governed, autonomous learning while taking courses (Posner, 2017). Adaptive learning (or teaching) is the real-time adaptation of instructional strategies, learning content, and educational resources based on the ongoing analysis of learner data. Integrated with AI, machine learning, and new data analytical techniques, these systems can now monitor the activities of learners, interpret their digital data following specific models that explain their actions, and continuously make micro-adjustments in real time to improve student’s performance and engagement levels (Posner, 2017). By seamlessly adjusting the sequence and difficulty of the learning tasks based on student’s abilities and preferences, these platforms create personalized content and custom learning experiences. They make learning more accessible by adapting to the learning strategies of individual students and altering and modifying the sequence and difficulty learners experience based on students’ demonstrated proficiencies (Zlatkovic et al., 2019). Educational and training organizations are increasingly adopting intelligent adaptive learning platforms as they attempt to scale the benefits of one-on-one personal tutoring to entire classes of learners.

 

Smart Adaptive Learning Platforms

Compared to lecture-based teaching, intelligent adaptive learning platforms are like using an interactive e-reader that allows for a personalized reading experience. The font size of the text and background color can be adjusted, and readers can be more engaged as they highlight, take notes, get instant definitions, and receive engaging and tailored recommendations based on their progress and interests. Accordingly, education organizations seek to emulate the advantages of an e-reader by adopting a wide variety of intelligent adaptive learning platforms to deliver recommended courses and customized learning experiences that align with individual academic interests or career development goals. For example, several universities (such as Northern Arizona University, Texas Tec University, and the University of Alabama) have adopted McGraw-Hill’s Open Learning Solutions as a blended learning approach that combines traditional face-to-face instruction with adaptive learning technologies. The types of existing systems include:

• Web3.0 Learning Management Systems (LMSs) (such as Blackboard and Canvas) can facilitate the seamless integration and data exchange between different educational platforms and dynamically adjust content by recommending additional resources or altering the sequence of course materials based on learner interactions and progress.

• Microlearning Platforms (such as Connecteam,  Whatfix, EdApp, iSpring Learn, and others) enable microlearning for employees by providing short and focused training content.

• Learner Engagement Platforms (such as Teachfloor, Orah, Google Workspace for Education, kahoot, and others) provide a collection of tools to help instructors, trainers, and course creators capture students’ inputs and track their performance to develop personalized learning experiences for their students (Teachfloor).

• Adaptive Learning Systems (such as Knewton Alta, Smart Sparrow, and SmartBook) provide authoring platforms for instructors and course designers, and assist universities or departments in developing a curriculum based on adaptive learning (Johanes & Lagerstrom, 2017).

 

User Interface to Track and Monitor Learners’ Progress

Most intelligent adaptive learning systems include a dashboard allowing learners and teachers to monitor and track their progress. Each customized interface is accessed through smartphones, computers, headsets, and other digital devices, as it manages the interactions and communications between students and the learning system. For example, data visualization tools (such as Tableau and Power BI) can transform complex data into intuitive visual formats and help learners and educators understand patterns and trends in educational data, making it easier to make informed decisions.

Selecting Adaptive Systems for Your Digital Hub
Because educational and training organizations require students to adopt adaptive learning systems apps for classes after selecting an academic organization to pursue educational and training opportunities, learners often need help deciding which apps to add to their digital hubs. However, it would be best to be keenly mindful that most systems currently providing smart adaptive learning opportunities do not use Block Chain technology to secure learners’ data. While adaptive systems offer vast and exciting opportunities for gathering valuable insights for both learners and educational providers, before enrolling in courses, you should be sure ethical considerations are in place to ensure your privacy and security in collecting and using your personal data.

Up Next: Blockchain Technology

The tremendous amounts of personal data generated from Web3.0 tools in adult learners’ digital toolkits should be protected. My next blog post will discuss the transition to blockchain technology to manage and secure students’ data.

 

Larry G. Martin, Ph.D.
Professor Emeritus, UWM
Follow me on X (formerly twitter) https://twitter.com/larry_martin29 and LinkedIn https://www.linkedin.com/in/larry-martin-142b528/

 

 

References

Bryant, J. (2024). College dropout rate in the U.S. https://www.bestcolleges.com/research/college-dropout-rate/#:~:text=In%202022%2C%2029.2%25%20of%20students,college%20experience%20but%20no%20degree.

Fischer, C., Pardos, Z. A., Baker, R. S., Williams, J. J., Smyth, P., Yu, R., ... & Warschauer, M. (2020). Mining big data in education: Affordances and challenges. Review of Research in Education, 44(1), 130-160.

Johanes, P., & Lagerstrom, L. (2017, June). Adaptive learning: The premise, promise, and pitfalls. In 2017 ASEE Annual Conference & Exposition.

Johnson, L., Adams Becker, S., Estrada, V., Freeman, A., Johnson, L., AdamsBecker, S., ... & Freeman, A. (2016). NMC Horizon Report: 2016 Higher Education Edition. Austin, Texas: The New Media Consortium. Retrieved August, 20, 2016.

Lee, L. K., Cheung, S. K., & Kwok, L. F. (2020). Learning analytics: Current trends and innovative practices. Journal of Computers in Education, 7, 1-6.

Peng, H., Ma, S., & Spector, J. M. (2019). Personalized adaptive learning: an emerging pedagogical approach enabled by a smart learning environment. Smart Learning Environments, 6(1), 1-14.

Posner, Z. (2017) What is Adaptive Learning Anyway? https://www.mheducation.com/news-insights/blog/what-is-adaptive-learning.html

Zlatkovic, D., Denic, N., Petrovic, M., Ilic, M., Khorami, M., Safa, A., ... & Vujičić, S. (2020). Analysis of adaptive e‐learning systems with adjustment of Felder‐Silverman model in a Moodle DLS. Computer Applications in Engineering Education, 28(4), 803-813.