Participants will learn to leverage the capabilities of machine learning to provide and monitor the implementation of targeted academic supports for students. These processes can be used to develop or enhance early alert systems by identifying struggling students and intervening before they experience a drop in academic achievement.
Online learning, while offering flexibility, presents a different experience than face to face learning. One area that is notably different is the physical absence of an instructor to provide just in time guidance and clarity for learners when they are struggling to understand a particular concept. Struggling students may suffer significant consequences from the lack of physical presence of an instructor. Instructors also face the challenge of identifying cues related to student performance.
The affective domain serves as the gatekeeper for learning (Tyng, Amin, Saad, and Malik, 2017). First time in college students, non-traditional students, or struggling students can experience this primitive brain function as a barrier to learning, especially in the online environment. Students, left to tackle difficult concepts, rigorous content, or even confusing directions, can easily find themselves at their frustration threshold, with neurological responses that will further prevent learning. This situation can be mitigated by providing struggling students with specific interventions to help them be successful, or, better yet, preventing these issues before students encounter them. Early, targeted interventions allow students to learn to their ultimate potential.
Designing early alert systems must be a deliberate and specific process. Ideal systems should include not only indications that students are in need of intervention, but details about what interventions would be appropriate to help the student to course correct. In this session, participants will learn how Artificial Intelligence (AI) can be applied in the online learning domain by analysis of quantitative indicators, as well as qualitative indicators, such as unstructured text submissions by teachers and students. Employing algorithms that understand both the quantitative indicators and the underlying sentiment of the submission, keywords, and semantic relationships, can crystalize the student story and help educators to better understand how to help learners reach their maximum potential. This understanding is critical to ensure that educators adapt teaching methods to individual students (Hanan, Hosam, & Fauzy, 2019).
In this session, participants will: 1) Understand the differences in online and face to face instruction. 2) Understand the importance of providing early interventions to students who are struggling in the online environment. 3) Understand the power of early alert systems that not only indicate students who struggle but also suggest appropriate interventions to course correct the student before they experience academic turbulence 4) Understand the role AI can play in the construction of effective early alert and intervention systems. 5) Engage with frameworks to begin the process of constructing or refining early alert and intervention systems.
To begin, participants will be assigned a perspective, through which they will experience a typical narrative of online coursework. Participants will share experiences from their perspectives and discuss potential barriers to learning students who experience these narratives would face. Participants will learn how these barriers can be circumvented by providing targeted interventions to students before they experience academic distress or academic failure. The role of AI can play in this process will be explored, with attention to creating or refining a system to include not only student identification but intervention recommendations. All participants will leave with a framework for building their own system, which can be applied to an online classroom setting.