2. What‘s in this session?
• Short overview of the field of Learning Analytics
• Why use Learning Analytics for SRL practice?
• Dealing with evidence in form of learner data
• What is needed to use Learning Analytics for SRL
(and in general)?
• Development of an EQF-mapped framework of
practical competences
• Your participation!
3. What is Learning Analytics
Basically, it is the use of „Big Data“ in education for the
sake of improving learning.
Definition:
“the application of analytic techniques to analyse educa-
tional data, including data about learner and teacher
activities, to identify patterns of behaviour and provide
actionable information to improve learning and learning-
related activities”
(Van Harmelen & Workman, 2012)
4. Why Learning Analytics?
Students (and teachers) appreciate:
- Better understanding of (own) learning behaviours
- Personalised assistance
- Timely feedback and intervention
- Information that facilitates independent learning
(Wong, 2017; Tsai, Scheffel & Gasevic, 2018)
5. Distinguishing features of
Learning Analytics
- Scale: large digital datasets (big data) from different
sources, micro-to-macro level information across entire
learner population
- Speed: LA provides automated real-time information,
on-demand timely feedback for learner positioning
- Sensors: Can collect data from variety of sources, incl.
different sensors (e.g. heart rate monitor, face
recognition, GPS, etc.)
- Authenticity: Collects behavioural data unobtrusively,
not influenced by „opinions“
7. What is learning evidence?
Objective information about learning process and
success (progress, achievements)
• Uses direct observation
• Uses scientific research from Learning Sciences
• Uses conclusive evidence (e.g. causality)
• Uses learning data information
• Authentic information (real setting, unobtrusive)
• Real-time information at micro and macro level
8. Learning Analytics for
evidence based learning
We see Learning Analytics as a reflection amplifier
and part of objective learning evidence that triggers
a learner and/or teacher response. This very strongly
links it to the self-evaluation strategy of the Zimmer-
man model.
9. Why Learning Analytics
for SRL?
SRL requires (regular) feedback during the learning
process
• To monitor progress and position
• To adjust self-regulation strategies
• To plan next steps
• To keep focus (goal orientation)
• To compare with others
• To identify and acquire successful learning strategies
• To motivate the learner
10. Why Learning Analytics
for SRL?
LA simultaneously provides different perspectives on
learning in context without interference
• Micro level (individual) – personalisation
• Meso level (class) – adaptation of teaching strategies,
identify outliers
• Macro level (all learners past and present, curriculum,
school) – school management, curriculum development,
pedagogic theory
11. SRL practice:
the Quantified Self
The „Quantified Self“ movement promotes self-
observation and reacting to one‘s own performance
data.
Mostly used for (self-selected) behavioural changes: quit
smoking, lose weight, get fit, sleep well,…
Learners take full charge of their own progress and
success. Reflection, context, and instruments are key.
e.g. fitbits, GPS watches, heart rate monitor, step
counter, mobile apps, fitness studio tools
12. Digital Education Action
Plan
European Commission (2018):
“Everyday exposure to digital data, driven largely by
inscrutable algorithms, creates clear risks and
requires more than ever critical thinking and the
ability to engage positively and competently in the
digital environment.”
13. Exploitation of learner data
for SRL
Three self-regulatory processes in SRL (Zimmerman, 1989):
• Self-observation
• Self-judgement
• Self-reaction
Data from learner behaviours can support these processes
BUT: using learning analytics requires a number of
competences
17. Data literacy
“Furthermore, the use of digital technologies in education, whether for
assessment, learning, administrative or other purposes, results in a wide
range of data being available on each individual learner’s learning
behaviour. Analysing and interpreting this data and using it to help make
decisions is becoming more and more important – complemented by the
analysis of conventional evidence on learner behaviour.”
(Redecker, 2017)
Seeing data in context!
20. Shortfall awareness
• Data and analytics are not neutral. Analytics are narratives. Data leads
to hypotheses (potential explanations) that need to be tested and
verified.
• Mostly come in probabilities not in clear answers.
• Pedagogic bias toward instructionism.
• Data not only shows learner information but implicitly also
teacher/context information.
• Cheating risks: gaming the system.
Requires critical thinking and holistic learner
perspective!
21. Transfering data to learning
SRL: Self-reaction based on self-judgement
or (guided by teacher): targeted intervention
3 areas of analytics-induced change in self-reaction:
22. Transfering data to learning
Mutual influences between learning evidence and
learning design.
Applying the right action after looking at the data evidence.
Reflect on learning design: what needs to change, how to
go about changing it? e.g. correcting misconceptions
Manipulating weighting of indicators: what is more
important to get a focussed understanding.
Requires: scientific thinking!
23. Ethical and legal knowledge
Requires transparency!
• Data protection (GDPR)
• Privacy protection: tracing of behaviours and activities
• Student surveillance vs. student support
• Obligation to act
• Data security
“Ethical and responsible data use is part of knowing how to
use data, and that knowledge focuses on how to protect
student privacy and maintain confidentiality of student data.”
(Mandinach, Parton, Gummer, & Anderson, 2015)
24. SRL data in a social context
Knowledge and skills are not solely individual
achievements, but are developed, carried forward,
and passed on through interaction and collaboration.
(Ferguson & Shum, 2012)
• Social aspects of learning: collaboration, roles, etc.
• Social positioning of the learner: performance
compared to peers, gamification
Requires data in social context!
25.
26. Exercise: Map your knowledge, skills and level of
autonomy on the EQF framework
Use marks 1 (low) - 5 (high) for each box
27. THANK YOU VERY MUCH !!!
Contact: wolfgang.greller@phwien.ac.at