I actually missed the live session of this talk, as it was scheduled at 11am MST instead of the usual 1pm. The email came in the morning during meetings, and by the time I tuned in, I had missed it. Thank you George for recording these sessions and making them available the next day!
Age of Social Media and Learning
Slidea of this prsentation
How do we know how well we are doing using any technologies? How successful are we in learning and how effective our institutions are?
Topics discussed/reactions and questions:
*Learning - today, tomorrow...
Today's idea of learning and general ed is a pizza box. Come to the institution, we will give you a textbook, study guide, instructor, and final exam. The final will be the info in the textbook.
Agree that this is not the idea of modern education that should include higher level learning, application, thought, analysis.
See Dragan's Learning Paths:
learning doesn't just happen in the box--formal and informal
idea of Educational Ecosystems
reusability
adaptivity
evolution
collaboration with educators and students
The authoring of these systems with reusability is it's packaging. to create
LEARNING AND COLLABORATION
mobile
administration
community
peer-review
presenting (youtube, recording and posting, blogging and discussing)
learners
students using many tools etc--still creating a bigger pizza box-within the boundaries of the institution. These BOUNDARIES ARE LIMITING WHAT WE WANT TO ACHIEVE.
How do we capture all the learning happening EVERYWHERE?
How can we connect all these tools in which students are linking with all the formal tools that students are using?
Having a formal or commonly shared model may be important--to integrate these systems
Three Generations of Distance Education Pedagogies
Connectivism
open idea-but boundary of institution that requires more formal course
border of connectivism and social constructivism
keep integrating
How can we facilitate learning analytics for constructivism and connectivistm -- while behaviorism is important--these had some analytics.
not to set back learning to behaviorism, but to learn about LA in ways to learn more about constructivism and connectivism
Part II: Learning Analytics
A systematic change is needed!
LA- What?
Measurement, collection, analysis, and reporting of data about learners and their contexts (SoLAR)
LA--Why?
Understanding and optimising learning and the environments in which learning occurs
and then how can we optomize the environment sna tools to make learning more effective
SoLAR paper about OpenLearningAnalytics
differentiation between learning analytics and academic analytics
LA benefits learners, educators, teaching staff, whereas academic analy benefits administrators, funders, marketing, governments, etc.
EVIDENCE BASED EDUCATION
As the integration of best research evidence with practioner expertise and stakeholder values (based on Evidence Based Medicine)
use the research and LA as a critical enabler to proceed toward evidence based ed
Criticism about LA-- concern that analytics will reduce the learning experience to equations if we return to behaviorism
LA involves many statkeholders not involved in pedagogical underpinnings of ed
accusations can be useful, we don't need behaviorism fears? because analytics can only be applied if we keep track of learners BEHAVIORS,
but this is more than just BEHAVIOR--one trace is action recording, go deep down and analyze blog text to see the cognitive presence of the person and measure critical thinking--machine algorithm analyzing the cognitive function and looking at constructivistm
When we look at the analysis and the social metric and person is involved in and their community, we are then talking about learning analytics which will support these more advanced pedagogies.
??-- To follow the clinical / medical analagy, how to perform analytics in a controlled manner? For example, software may differ among so many dimensions, e.g. UI, UX, content, learning paradigm employed, what is going on in the classroom that is prompted by the software. Where a given material subject to analytics may be aiding learning along some of these dimensions, it may be harming along others. The studies of learning materials I have seen don’t even beging to adress these issues and tends reduce results to a binary ‘the software improves / does not improve learning’.
Part II
Learning Analytics
Approaches and Challenges
with bits of semantics
BIG DATA-- need machines, data overload, we are producing more content on the web than we are ever able to keep up on--now connectivism- ? is how can we deal with these?
LA CHALLENGE: WHAT AND HOW TO COLLECT?
Ubiquitous Learning analytics?--Tool and format independent--aggregates and integrates
how can we gather data from ALL formats, and aggregate to give feedback to learner and others in the stakeholder process
SEMANTIC WEB
Ontologies-- interconnecting application--shared domain conceptualizations
Peter Metcalf paper-- ontologies
Linked Data Graphic
DPedia-- rdf or other techs --all the underlying structures and concepts in wikipedia extracted and processed by machines to easily navigate from one concept to another concept--
therefore, when we are analyzing blogs in LA, we may be
"A crazy problem requires a crazy solution!" --Griff Richards, 2005
NEED AN ONTOLOGY TO CAPTURE ALL THESE PEDAGOGICAL ISSUES
Learning Context Ontology: LOCO captures many kinds of data:
content structure ontology
content type ontology
user model ontology
learning design ontology
domain ontology
Jovanovic, J., Knight, C., Gasevic, D., Richard, G., "Learning Object Context on the Semantic web." 6th IEEE International Conference on Advanced Learning Technologies, Kerkrade, The Netherlands, 2006 pp.669-673.
good paper for ontologies
http://www-ksl.stanford.edu/people/dlm/papers/ontologies-come-of-age-mit-press-%28with-citation%29.htm
another good paper on ontologies - general overview: http://www.jamesodell.com/Ontology_White_Paper_2011-07-15.pdf
EDUCATION SOCIAL SEMANTIC WEB
Combing the Social and Semantic Web
Formative Evaluation
Ali, L., Hatala, M., Gasevic, D., Jovanovic, J. (2012). A Qualitative Evaluation of Evolution of a Learning Analytics Tool. Computers and Education, 58(1) 470-489, http://goo.gl/ICvMT
LA HOW AND WHAT TO REPORT?
Visual Learning Analytics
Ubiquitious LA impossible without visual because of information overload
how we can inform our educators better to different conditions students are demonstrating and compare students to each other--overload?
provide ed insights about cognitive functions
wanted to compare types of social activites--kinds of posts, different kinds of conceptualizations being discussed
types of sociographs related to different concepts
allow students to do social bookmarking and highlighting, and see what concepts students
Learning Analytics Acceptance Model graphic
Inspired by Venkatesh, V., Morris. M. G., Davis, G. B., & Davis, F. D., 2003 User acceptance of information technology: Toward a unified view. MIS Quarterly, (27:3), pp.425-478.
a model to predict learner potential
analyze different LA related to quality of concepts (course, modules, class), with types of student interactions (engagement in lessons, discussions, general comprehension), see how all are associated with LA and ease of use--study which provided and identified correlations-- one critical points about general student interactions is strong association of social analytics--most important is the social interaction analytics that lead to the overall perceived usefulness of the analytics and the tool.
The adoption of a tool is based on the context, policies, cultural,
LA WHAT TO MEASURE?
Measurement, collection, analysis, and reporting of data about learners and their contexts
How can we increase level of critical thinking? community of inquiry used as a measure of critical thinking--
cognitive level can be associated with different kinds of teaching, social presence,
LEARNING ANALYTICS FOR COMMUNITY OF INQUIRY
EFFECTS OF INSTRUCTIONAL INTERVENTIONS
Example: Role playing (invited expert and moderation) with explicit instructions how to contribute.
how do I maintain and identify that this interaction works?
How can we identify different kinds of metacognition?
Social Learning analytics for self-regulated workplace learning--LearnB-- for each
another paper Dragan
for differnt types of artifacts, we want to provide different streams--if I'm studying something, don't show me something distantly related
what kinds of events are around certain competencies? how to facilitate with visualization with the level of peer knowledge, if I created a new competence and shared it with everyone, I can see my performance and sharing with the rest of my peers--only me to see this--I can regulate my own behavior
MEASURING COGNITIVE PRESENCE
TEXT MINING AND LINKED DATA
A very similar text-mining problem is spam classification.
How id cog pre AUTOmatically
by analysis of the text--.57-.74
we have anti spam 95% or higher
these are the kinds of algorithms to auto classify msgs based on cognitive presence
COGNITION AND META-COGNITION
Discovering learning processes
to see indiviudal numbers, but see the flow of activities, first msgs in infor sharing, then usually see resolution cog presence
area of process mining
chart flows--
SNAPP
http://goo.gl/jtO31
http://nodexl.codeplex.com/
Technology Acceptance Model as a model for the above
Learning Analytics
MORE
meaningful and ubiquitous
with semantic technologies
to label and not do spam detection or labeling--do text analysis and determine
triggering event
exploration
inspiration
resolution
other
No comments:
Post a Comment