3. What is Learning Analytics?
The measurement, collection, analysis and
reporting of data about learners and their
contexts, for purposes of understanding and
optimizing learning and the environments in
which it occurs
(Society for Learning Analytics Research – SoLAR)
5. Learning Analytics Community
Exchange (LACE project)
● European Union Support & Coordination project
● 24 more months to go
● 9 partners (Norway, Sweden, UK, Belgium,
Netherlands and Italy)
● Focus areas
o Schools
o Higher Education
o Industry
o Interoperability & Data Sharing
6. ISO.. SC36 WG4 Study Period
Studies:
● Korean report on Big Data & LA
● IMS Caliper
Proposal:
● Should SC36 develop a multi-part standard?
● What would be the sources?
7. Possible parts in a new ISO standard
● Conceptual Framework
● System Requirements
● Privacy Protection Guidelines
● Accessibility Guidelines
● Quality Metrics
● Data Crawling
● Reference Software
● Implementation Guide
8. ADL et al.
xAPI
● Basic data model:
o ‘Learner’ ‘created’ ‘activity’
● High level property vocabulary
● An API for communicating activity
● Open source implementation
9. IMS Global
Caliper:
● Vocabulary for student activities:
o “metric profiles”
● An API for communicating activity data
o “Sensor API”
● Designed to work with existing IMS specs:
LTI, QTI
● Implementation imminent
10. Open Learning Analytics (SoLAR) initiative
● Summit April 2014 (after LAK 14)
● Identified OLA domains:
o open research (e.g. open datasets, open predictive
models, etc.),
o institutional strategy and policy issues, and
o learning sciences/learning design and open
standards/open-source software
● Open Learning Analytics Architecture
11. LACE Interoperability work so far...
● Started a series of Webinars
● Identified that 'soft issues' are extremely
important from an implementation
perspective
● Consensus needed on privacy issues,
access to data, users’ control of their own
data, stakeholder responsibilities, etc.
12. CETIS conference cultural issues
● How do we convince management of the return on investment of
analytics?
● Where do we start: the problem to be addressed or the data available?
● What information do we not want to know? (privacy law)
● How do we get out of the deficit model of student achievement?
● How do we overcome resistance to change?
● Where do we get statisticians / data scientists from?
● How do we make insights actionable?
● How do we deal with the US bias in the data structures of some of our
systems?
13. CETIS conference tech issues
● How do we distinguish correlation from causation?
● How do we get reliable data / avoid students gaming the system?
● How do we identify the useful data in the wealth of available data?
● How do we make sense of raw data; what does "student is in the
building” mean?
● How do we get relevant data from outside the institution?
● How do we get data from multiple (legacy) systems to
interoperate?
● How do we deal with missing longitudinal data?
● How do we make data comparable between institutions?
14. Workshop on LA in Japan Nov/Dec
1st ICCE workshop on Learning Analytics (LA):
Leveraging Educational Data for Adaptive
Learning and Teaching
ICCE 2014 conference in Nara (30 November -
4 December)
Submit your workshop paper!
15. And then we have MOOCs...
This Monday we got the Norwegian MOOCs
Commission’s Report
● Suggests 2 mill € yearly for Research,
Development and Community Exchange
Activities the next few years