Hello everyone, this is Geng from D3. In this article, I would like to share with you the paper we discussed in our recent English literature seminar and my impressions of it.

Paper Title: Developing a Learning Analytics Intervention in E‐learning to Enhance Students’ Learning Performance: A Case Study
Journal: Education and Information Technologies
Volume and Pages: Vol. 27, pp. 7099-7134
Publication Year: 2022
Authors: Si Na Kew, Zaidatun Tasir

Now, let me provide you with a summary of the information presented in the paper.

In recent years, universities have increasingly embraced e-learning. In this mode of learning, students progress at their individual pace and knowledge levels, leading to variations in their learning trajectories. Consequently, there arises a necessity in e-learning to tailor it to the unique needs of each student, to pinpoint challenges they encounter, and to offer support for their learning endeavors. Moreover, student motivation assumes a pivotal role in the realm of e-learning. Specifically, when students lack motivation, they exhibit reduced enthusiasm for learning and are more prone to dropping out of a course. Student cognitive engagement and knowledge retention also pose challenges within the e-learning landscape.

Learning analytics (LA), as referred to hereafter, presents a solution to the aforementioned challenges in e-learning by scrutinizing student learning behaviors and implementing interventions in the learning environment to enhance the learning experience. Numerous universities, including Purdue University and Northern Arizona University, have employed LA to devise interventions aimed at supporting student learning. However, these educational practices suggest that interventions solely targeting learning objects and materials in e-learning might prove insufficient when not aligned with students’ diverse learning styles and motivations.

Hence, this study undertook the development of an LA intervention that integrates the Felder-Silberman learning style model and the ARCS model. The investigation focused on assessing the effectiveness of this LA intervention concerning motivation, learning outcomes, cognitive engagement, and retention in the context of e-learning.

The authors of the paper conducted an evaluation of a 17-week Learning Analytics (LA) intervention involving 50 sophomore undergraduates enrolled in a computer-based course. The study was executed through the following steps:

– Weeks 1 through 7: During this period, students had unrestricted access to learning objects via eLearning. The data generated by the eLearning platform were utilized and analyzed to identify students at risk based on factors such as motivation, engagement, knowledge retention, and learning outcomes. Decision tree techniques in data mining were employed for this analysis. Additionally, students identified as at-risk were categorized based on their learning style data.

– Week 8: The LA-based intervention was implemented during this week. A new learning object was created by amalgamating the learning styles model and the ARCS model. Motivation and learning style factors were incorporated according to the students’ class themes. Learning data generated through the e-learning program were collected during this phase.

– Weeks 15 to 17: In the final weeks of the study, the collected learning data were subjected to further analysis. This analysis aimed to measure student motivation, learning outcomes, cognitive engagement, and knowledge retention, providing an assessment of the overall effectiveness of the LA intervention.

For the analysis of learning logs and psychological data, student motivation was gauged using the IMMS (Instructional Material Motivational Survey) questionnaire both before and after learning sessions. Data obtained from e-learning discussions were employed to compute student cognitive engagement, with multiple raters evaluating the data based on the coding scheme proposed by Van der Meijden (2005). The assessment of students’ learning styles was conducted using the Index of Learning Styles (ILS) learning style diagnostic questionnaire.

Furthermore, learning outcomes and knowledge retention were appraised through pre- and post-tests that were aligned with the course content. These comprehensive assessments allowed for a thorough examination of the impact of the LA intervention on various facets of student learning, encompassing motivation, cognitive engagement, learning styles, and academic achievement.

The results obtained through Wilcoxon’s signed rank test revealed a statistically significant increase in both student motivation and knowledge retention following the implementation of the LA intervention. Additionally, t-tests conducted on the pre-test and post-test scores demonstrated a statistically significant difference, indicating that the LA intervention positively influenced learning outcomes for students who underwent the intervention.

Moreover, in terms of cognitive engagement, there was a notable increase in the number of high-engagement students, rising from 16 to 48, while the number of low-engagement students decreased from 17 to 2 after the intervention. This implies that the LA intervention effectively heightened students’ cognitive engagement. In essence, the LA intervention played a significant role in enhancing students’ motivation, learning outcomes, cognitive engagement, and knowledge retention.

This underscores the importance of early identification of students at risk of learning delays in e-learning and tailoring learning objects based on their individual learning styles. The findings suggest that LA interventions providing personalized learning objects, aligned with students’ needs, can positively impact learning outcomes, motivation, and engagement. Specifically, it is anticipated that when LA interventions deliver learning objects tailored to students’ explicit content structure and learning style, students will efficiently complete e-learning assignments and acquire knowledge. Furthermore, the integration of motivational components from the ARCS model into learning objects may boost student motivation and participation in discussions, facilitating easier knowledge retention.

Here are my reflections on the paper. Learning Analytics (LA) extends beyond mere analysis; its primary objective is to assist learners in comprehending their learning processes, facilitating improvement, and ultimately enhancing the learning experience. LA-based interventions have been promoted as effective tools for improving learning, and I am particularly intrigued by their approach, which is why I chose this paper. I have previously delved into research on delivering learning content in e-learning using the Felder-Silverman learning style model. Still, I found this study to be innovative and informative due to its integration of the learning style model and the ARCS model for delivering learning content.

However, I am keen to understand how LA interventions can be effectively provided for at-risk students, as this aspect is not well-elaborated in the original source. For instance, I wonder whether providing active-type learning objects that align with students having active learning styles or opting for learning objects of a different style, such as introspection type, would be more beneficial. The manner in which these learning styles are presented should also be considered a crucial factor for the intervention’s success.

Furthermore, the current analysis results suggest challenges in identifying causal relationships between learning outcomes, cognitive engagement, and motivation. Employing additional analytical methods could help elucidate these causal relationships, leading to more intriguing findings.