WILSON, Richard S. (2013). Using business intelligence to predict student behaviour. Doctoral, Sheffield Hallam University (United Kingdom).. [Thesis]
Documents
20551:489165
PDF (Version of Record)
10701198.pdf - Accepted Version
Available under License All rights reserved.
10701198.pdf - Accepted Version
Available under License All rights reserved.
Download (65MB) | Preview
Abstract
In light of reduced Higher Education funding, increased student contributions and competition between institutions, finding ways to understand student progression and improve the student experience are integral to the student, institution and state (York and Longden 2008). This research uses Business Intelligence, specifically Data Warehousing and Data Mining, to build models that can be used to predict student behaviour. These models relate to final award classification, progression onto postgraduate studies at Sheffield Hallam University and employment type post undergraduate degree completion. This work builds upon the recommendations of Burley (2007) where the Department of Computing, at Sheffield Hallam University, was used to prove the applicability of such techniques. It is fair to state that the field of student progression has been well documented over the years. Numerous authors (Tinto 1993, Yorke 1999, McGivney 2003) have all developed strategies and intervention techniques to help aid student progression. The evolving field of Educational Data Mining has focused, in the main, upon student interactions with web-based learning environments (Romero and Ventura 2006). Few studies have tackled the subject of using Business Intelligence as a method of understanding student progression (Dekker et al 2009, Herzog 2006). The data was collected from the universities information systems and through the process of Data Warehousing and Data Mining a number of predictive models were constructed. This resulted in the identification of some interesting rules and variables, such as course and ethnicity, which are also fundamental in the more traditional student progression literature, such as Yoke and Longden (2008). Overall, this research has further proved the applicability of Data Mining in Higher Education. The major institutional findings that have been established are: added value students are more likely to take postgraduate studies at Sheffield Hallam University, and a student's ethnicity can influence progression onto postgraduate studies and obtaining a graduate job.
More Information
Statistics
Downloads
Downloads per month over past year
Share
Actions (login required)
View Item |