Learning analytics is themeasurement, collection, analysis andreporting of data about learners and theircontexts, for purposes of understanding andoptimizing learning and the environments inwhich it occurs. Society of Learning Analytics Research
“A university where staff and studentsunderstand data and, regardless of its volumeand diversity, can use it and reuse it, store andcurate it, apply and develop the analytical toolsto interpret it.”
Roots of learning analytics Statistical methods Intelligent EDM Tutors Big Data Personalization Business Learning AdaptiveIntelligence learning Analytics
We can make great progress with analytics asindividuals, but the nature of ‘big data’ requiresa systemic focus
SNA: in Moodle introduction forumLimitedinteraction.Most areisolated
Source: King James Bible Old and New Testaments
Tools for analytics- Single functionality tools (SNAPP)- Multi-function tools (ManyEyes, Gephi)- Visualization (Tableau)- Existing research tools (SPSS, Wolfram)
Acquisition: how do we get the data – structuredand unstructured?Storage: how do we store large quantities?Cleaning: how do we get the data in a workingformatIntegration: How do we “harmonize” varying datasets togetherAnalysis: which tools and methods should be used?Representation/visualization: tools and methods tocommunicate important ideas
Many tools are currently stand alone and singlefunctionality(anyone remember learning managementsystems in the late 1990’s?)
PlayersStartups (Grockit, Knewton)Existing vendors (Cisco, Catatel)Systemic learning analytics players (iStrategy)(they get bought up quickly)
Check my activity Predictive Analytics Reporting
To date, asystem-wide open source analytics platformis NOT available.We want to change that.
Given how important analyticsare…maybe we should start with openness
The Vision:Open Learning Analytics Architecture
OLA: An open, extensible platform for researchers, educators, and administrators for broad-scale analytics in education and learning.
Principles of a systems-wide analytics tool1. Algorithms should be open, customizable forcontext2. Students should see what the organization sees3. Analytics engine as a platform: open for allresearchers and organizations to build on4. Specific analytics strategies and tools: APIs5. Integrate and connect with existing open tools6. Modularized and extensible