Exploring the Potential of Learning Analytics to Measure Student Participation and Engagement: Researchers’ Experiences from an Exploratory Study

Year: 2018

Author: Fan, Si, Garg, Saurabh, Kregor, Gerry, Yeom, Soonja, Yanjun, Wang

Type of paper: Abstract refereed

Abstract:
Big Data analytics is commonly used to inform decisions in different areas, including higher education. The most common form of Big Data analytics in higher education is Learning Analytics, which is a relatively new concept to capture educational Big Data with a purpose to enhance the student learning experience. This research was conducted at one Australian regional university. The project aimed to examine: (i) the relationship between students’ interactions with the unit contents and other users, and their assessment results; and (ii) the influence of teaching staff behaviour on students’ engagement and assessment results.
Data were extracted from seven units in two disciplines: Education and Information and Computer Technology (ICT). The data were collected through the Learning Management System (LMS) used at this university. All the chosen units had an online component, with six of these units being blended units, involving both face-to-face and online delivery, and one unit being purely online. The data include the number of times the students and teaching staff interacted with: (i) the unit content delivered through the LMS: lectures, learning materials, quizzes, etc.; and (ii) other students and teaching staff in the unit, as shown in the number of discussion board postings.  
Initial findings identified important factors that significantly affect students’ engagement in the units. For example, to promote student engagement, increasing the interactions, including discussions, between teaching staff and students, would be more effective than modifying the unit content. Also, the tutors had much less online interactions with students than the lecturers, despite the known importance of tutor input in promoting students’ engagement.
This research faced a number of challenges. First, the quality and validity of the data collected from the different units was largely uneven, due to the different unit design and presentation. In addition, the current collected data can only indicate the users’ behaviour to some extent. Students’ usage of the LMS can be indicated by their time spent in the system, however, there are risks in using the total length of time spent as the only indicator for student engagement. Further data collection and analysis should include the number of clicks performed by users during different time periods in a day, and discussion contents, which could reveal students’ level of understanding of the learned knowledge.
Acknowledgement
The project was supported by the University of Tasmania, Australia, under the Research Enhancement Grant Scheme (REGS).

Back