Sunday, November 23, 2014
Ascilite 2014 workshop on learning analytics
Learning analytics workshop with Dr. Leah Macfadyen and Dr. Shane Dawson at Ascilite 2014 held at Otago Polytechnic.
An introductory session. Leah is from university of British Columbia and Shane is from university of South Australia. Handouts from http://bit.ly/11HnjmH
Overall a good overview acknowledging challenges and potentials.
Unedited notes which I will augment when I return to office.
Current LA implementation
1 extraction and reporting of transactions level data
2 analysis and monitoring of operational performance
3 what if decision support - such as scenario building
4 predictive modelling and simulation
5 automatic triggers and alerts - interventions (Goldstein and Katz, 2005)
Most institutions have not progressed beyond stage 1.
Introduce and review field
Cooper 2012 " the process of developing actionable insights through problem definition and the application of statistical analysis.
Solar - measurement collection analysis and reporting of data about learners and their contexts for purposes of understanding and optimising learning and the environments in which learning occurs.
La draws from business intelligence, web analytics, academic analytics fro 2005, educational data mining 2000 , action analytics 2008 onwards.
Academic analytics on recruitment, retention etc. learning analytics focused more on teaching and learning.
Drivers of la include socio political, educational and technological.
Need to move beyond number crunching. EDM laid groundwork for computation and modelling. Interdisciplinary collaboration required technical analytic on statistics, data visualisation and visual analytics, educational data mining, computer science, machine learning, natural language processing, human computer design, etc.
Systems thinking may be one way to look at la. A system is perceived as elements which are inter connected and inter dependent ISON in Ray and bradbuRy, 2008.
LA investigates information of educational system information flows and feedbacks in an education system.
What kind of information? Where students come from, achievement of students, information from student engagement through LMS or similar MOOCs, student monitoring and success tracked by learners, student satisfaction from course evaluations etc.,
Student led system to feedback to students progress through courses. Gathers student demographics.
University of Maryland Baltimore 'check my activity' tool.
Austin Peay university's degree compass tool. More of an academic analytic looking at students pass achievements and recommends future courses.
University of South australia - analytics dashboard. Student learn online engagement ratio. Mainly frequency of access to various Moodle resources and activities. Then calculate at risk on quadrant student current engagement vs student current grade. Low engagement often related to low grades. Link to predicting performance and retention through identifying no risk, low, medium and high. Found intervention with high risk students too late, so shifted intervention to earlier in course, just post orientation. Risks identified as low social economic status, low entry qual. And distance to travel to uni. Students contacted via crm which is used by student services to support students.
Shane also shared uniSA work on video analytics and annotations. Collaborative lecture annotation system, allows YouTube video to be annotated with student notes.
Other enterprise level la tools include cavitas, design 2 learn insight, knewton adaptive learning system, blackboard analytics.
Barriers to adoption (ferguson et al., 2004)
Review key steps in implementation
Rapid outcomes model approach
Define and redefines your policy objectives
Map political framework (young and mandizabal, 2009)
Identify key stakeholders
Identify desired behavioural changes
Develop engagement strategy
Analyse internal capacity to effect change
Establish monitoring and learning frameworks
Define a clear set of overarching policy objectives
Map the policy context
Identify key stakeholders
Identify desired changes outcomes
Develop a plan
Ensure engagement team has competencies required to operationise
Establish monitoring or evaluation
What does your institution would like to achieve through implementing LA?
Clarify and prioritise purpose
Plus identify range of skills systems and processes to assist implementation.
Issues with ethics - purpose, ownership, informed consent, privacy, de-identification, how data handled, who has access, ethics of surveillance, but in parallel with changing attitudes to privacy and self disclosure. Need policies on data governance.
Workshop closed with session on ethics.
Issues with ethics - purpose, ownership, informed consent, privacy, de-identification, how data handled, who has access, ethics of surveillance, but in parallel with changing attitudes to privacy and self disclosure. Need policies on data governance. Recommended to check slade and prinsloo 2013 for principles of learning analytics and ethical use of student data for la policy (2014) by uk open university.