The annual Learning Analytic & Knowledge (LAK) conferences commenced in 2011 and have contributed significant research to this area.Papers submitted to the conferences,together withEducause Learning Initiative,provided invaluable sources for literature. Alfred Essa’s explanation of the LA tools available resonate with me in that he describes the tools we build for LA as being “akin to telescopes or microscopes that will hopefully allow us to see deeper, wider into student learning” thus optimising the student experience(Chalex, Essa, & Norris, 2012) .
The literature provides various examples of HEI applying data mining techniques. A numberof HEI in the United States are amongst the earliest leader adopters of LA. These include the Signals Project in Purdue University (Arnold, 2010), the establishment of the Performance Analysis Framework(Wagner & Ice, 2012) ,the early warning systems adopted by Paul Smith’s College in the US (Taylor & McAleese, 2012) and University of Michigan (Lonn, Krumm, Waddington, & Teasley, 2012). To narrow the scope of this study, the criteria for literature review focused on research papers adopting mining techniques on LMS data.
Case studies conducted by HEI have focused on student retention and identifing at-risk students. In an effort to address increasing demands for student success and institutional accountability HEI have witnessed a surge of interest in data mining and learning analytic technology in order to identify at-risk students and increase student retention rates(Campbell & Oblinger, 2007, pp. 3-6 ;DeBlois, Campbell, & Oblinger, 2007, p. 42).
During my literature review, I found research conducted at the University of British Columbia to be a significant research project(Macfadyen & Dawson, 2010). Similarites exist between the research carried out in their project and the research proposed in this document. They mined Blackboard data and utilised the SNAPP tool to carry out social network analysis to identify student/student and student/lecturer communication. They investigated which student online activities accurately predict academic achievement and which LMS data variables represented student effort or activity. Their research concluded strong correlation exists between LMS activity and course outcome particularly with discussion forum activity. This is further evidenced by Smith, Lange, & Huston (2012, p. 60) ;Dawson, Mc William, & Tan (2008, p. 227) whose findings suggest active site engagement with LMS can serve as an effective predictor of course outcomes. It was interesting to note fromDawson et al. (2008, p. 224), when referring to web 2.0 technologies, how mining of LMS data can provide valuable insight about lecturers pedagogical practice as they contend that data gleaned from LMS can be used to “guide and inform the diffusion of technology and integration into learning and teaching activities.” Such findings enlighten lecturers of the technology tools most effective in their teaching practice.
The readings highlight a concern relating to the utilisation of LMS that may have implications for this proposed study. In studies carried out by(Black, Dawson, & Priem, 2008 ;Dawson, Mc William, & Tan, 2008 ;Macfadyen & Dawson, 2010)LMS tools i.e. forums, wikis and discussion forums were used by students and lecturers. Since commencing research on LA, I have asked academic staff to what extent Blackboard is used and for what purpose. Feedback from informal discussions suggest LMS is regarded as a data repository. Previous research exploring LMS provided further evidence of this finding (Dyckhoff, Zielke, Bültmann, Chatti, & Schroeder, 2012, p. 58). Criteria for selection of DIT class codes to apply this research to will be based on the usage of web 2.0 tools within Blackboard in order to circumvent the practice.
Notably (Educause, 2011, p. 2) claimsstudents and instructors may feel that LA “takes assessment out of the realm of human judgment and reduces it to numbers and statistics”. The literature review uncovered similar findings withCampbell & Oblinger (2007, p. 16) raising the issue when highlighting concerns for HEI implementing analytic projects, contending such projects may be perceived as “dehumanising the educational process”. Diaz & Brown (2012) claim despite the benefits of LA technology, it should not be used by academics to abrogate their responsibilities in decision making as educators. Its role is to support sensemaking not supplant it(Brown & Diaz, 2012, p. 3). My thoughts on mining data are to inform the necessary stakeholders.It should provide intelligent data to enable better decision making. Whilst I am keen to explore this area, the proposal may be perceived as advocating the adoption of LA in HEI. I stress the research merely proposes an exploratory study to see beyond the “fog” of LA and determine to what extentits findings can be utilised. In line with thisSiemens contends
in spite of the potential of LA , it is important to emphasise that LA is not the end goal, it serves to provide stakeholders with better information and deep insight into the factors within the learning process that contribute to learning success. (Siemens, et al., 2011, p. 5)
Whilst LA offers much promise, the literature review highlightsa major concern namely data privacy. This has been raised by a number of experts in the field. Understandably students mayperceive LA as an invasion of privacy as the monitoring and tracking of student online engagement raises the “specter of a digital big brother”(Educause, 2011, p. 2). The concern was highlighted at the first LAK conference in 2011 at which attendees were in agreement that LA “raises deep and complex privacy issues” (Brown M. , 2013, p. 3) . This is one of the major challenges faced by LA that I identify with in the context of theproposal. I need to tactfullystrike a balance between learner privacy and the value of data collection for improving learning.
