[WORKSHOP] Sasha Poquet & Srećko Joksimović
This hands-on tutorial will focus on understanding networks. The participants will be introduced to the main concepts in network analysis, and will learn to analyse real-life networks with the tool of participants’ choice (Gephi or RStudio). [KEYNOTE] Associate Professor Carolyn Rosé
An important research problem in learning analytics is to expedite the cycle of data leading to the analysis of student needs and the improvement of student support. On the basis of the importance of social interaction in learning,in this talk I propose a pipeline that includes data infrastructure for a common representation of social interaction data from multiple platforms; a probabilistic sequence model to analyze the effects of social connections on students’ learning paths; and a social recommender system to support students for acquiring social connections with the potential for positive impact.The foundation of computational analytic work is representation of data. Once the data of interest has been represented in a way that is generalizable across sources, the next step is to model student trajectories, especially as they relate to their observed social connections. I will explain how this analytic approach enables us to identify opportunities where interventions can positively impact student trajectories. We propose a model that automatically extracts student learning paths composed of discussions across multiple platforms and active social engagement. This model aims to detect the pattern of students’ learning paths conditioned on their social situation and thereby inform us of the influence of different configurations within the social space. I will describe a recent application of this modeling approach to a data set from a cMOOC where the model revealed an important problem. If students followed other students who had set goals for the course, they were more likely to stay longer in the course, engage in hands-on practice of software tools, and link different materials they learned across the course. However, the majority of students were found not to take advantage of social connections in their course participation, and many students who chose to follow other students, chose to follow students who were not good role models with respect to goal setting. Thus, our solution involved an analytics enabled social recommendation approach to propose followees that would be natural choices for the students they were recommended to while also being excellent role models. |
[PRESENTATION] Catherine Spann
A common goal within higher education is to create purposeful and fulfilling lives. The increased use of digital technologies, and related impact on attention and focus, intensifies the challenges of achieving this mission. The concerns of fast-paced lives and distractive technologies have contributed to an explosion of interest in secular mindfulness strategies. Over the past several decades there has been a dramatic rise in studies and nationally funded research grants on the use of mindfulness practices. This presentation will review the theory and research on mindfulness and its relationship with self-regulation and psychological well-being. Furthermore, this talk will describe mindfulness in the context of teaching and learning and the ways in which technology can both enhance and interfere with these practices. [KEYNOTE] Professor Peter Reimann
Learning Analytics has the potential to contribute to research on learning: to deepen our understanding of learning processes, and to advance the learning sciences. For its potential to come to fruition, LA has to move beyond an event- and activity-centric ontology of learning to one that is informed by a view of learning as a complex phenomenon with emergent properties. I will illustrate these points by examples of using process-mining methods in the context of learning in small groups and for analysing data on self-guided learning. [PANEL] EASS Teaching Academics Network panel discussion with Dr Ruth Fazakerley and Vita Kovanovic
|
[KEYNOTE] Professor Simon Buckingham Shum – Teaching & Learning Breakfast Series
As algorithms pervade societal life, they are moving from the preserve of computer science to becoming the object of far wider academic and media attention. Many are now asking how the behaviour of algorithms can be made “accountable”. But why are they “opaque” and to whom? As this vital discussion unfolds in relation to Big Data in general, the Learning Analytics community must articulate what would count as meaningful questions and satisfactory answers in educational contexts. In this talk, I propose different lenses that we can bring to bear on a given learning analytics tool, to ask what it would mean for it to be accountable, and to whom. It turns out that algorithmic accountability may be the wrong focus. |
[PANEL] Professor George Siemens, T. Lisa Berry and Justin T. Dellinger
|
[PRESENTATION] Nia Dowell
Our language and discourse can provide insights into our social, cognitive and affective processes. My research program falls under the banner of a new field called computational discourse science. This new interdisciplinary field encompasses a broad cross-section of topics in cognitive science, psychology, education, and other areas of the social sciences that analyze language, discourse, and text. The scientific goal is to develop and test models of communication and computer-mediated learning. My research integrates computational linguistics with psychological theories of social interaction, discourse comprehension, and communication. This talk will present results from recent projects, which highlight the advantages of an interdisciplinary approach that uses language to assess knowledge construction and social dynamics during collaborative interactions. |
[KEYNOTE] Professor Dragan Gasevic
This talk will describe underlying principles, design, and experience gained with ProSolo, a platform that supports personalized, competency-based learning through social interaction. Traditional educational models are primarily focused on classroom education and training typically associated with the notion of credit hours as the (only) route towards formal credentials. This limits opportunities for creating personalized learning pathways in the changing educational context. ProSolo provide users with the ability to unbundle education programs, courses, and units into discrete yet inter-related competencies, allowing learners to construct their education pathway in a manner that better reflects their interests and future career motivations and requirements. ProSolo is developed with the intention of providing learners with opportunities to customize, modify, and personalize their self-directed learning journey. ProSolo supports the development of skills for self-directed learning by allowing learners to control the planning, learning, and presentation of outcomes associated with their learning. To support learners with different levels of prior knowledge, study skills, and cultural backgrounds, ProSolo offers features for supporting self-directed learning through three types of scaffolds, including instructional, social, and technological. Learning in ProSolo occurs within a socially rich environment that aggregates learners’ information created and shared in their existing online spaces. ProSolo makes use of learning analytics to empower learners and instructors in this new model of education. ProSolo was used in the Data, Learning, and Analytics MOOC and is currently being piloted at several university sites. |
[PRESENTATION] Dr Abelardo Pardo
Learning analytics emerged as a research area that uses data to gain insight on the learning process. Initial efforts focused on detecting students at risk and provide interventions to reduce attrition. Other relevant initiatives are studying the use of visualisations, student interactions in forums, written assignments, etc. But, is it possible to assist students with their overall learning strategy? Can data and algorithms provide instructors with resources to provide student advice at the course level and not at the activity level? In this talk we will explore the challenges to go beyond detailed clickstream data to higher level feedback constructs. |
[PRESENTATION] Associate Professor Jelena Jovanovic – Teaching & Learning Breakfast Series
The impact of active learning on educational outcomes has proved to be both large and consistent. Flipped/inverted classroom is an active learning strategy, which has been gaining increasing acceptance in a variety of disciplines. Likewise, researchers are providing more and more evidence for the benefits of the flipped learning approach. In particular, the literature reports on students’ positive opinion on and satisfaction with the flipped learning, as well as the improvement in students’ performance as measured on pre- and post-tests and course exams. However, the learning strategies that students apply in this novel learning framework have scarcely been examined. Considering that the flipped classroom (FC) model fosters student ownership of learning, and is quite different from the ‘traditional’ lecture model, it would be interesting and important to shed some light on how students approach and manage this new learning context, and how they organize their learning. Additional motivation for examining students’ learning strategies lies in the fact that it is not sufficient that students develop ‘only’ discipline-specific knowledge and skills during their formal education; it is also important, even essential that they learn to regulated their learning, i.e., to develop skills of contemplating, adapting and continuously improving their learning strategies. Therefore, in this talk I will present a method for examining students’ learning strategies in a FC, and illustrate its application in a FC-based engineering course. The method relies on the trace data collected during students’ interaction with the course content, and makes use of different analytic techniques – Sequence analysis, Clustering and Hidden Markov models – to detect patterns in students’ learning activities. This exploratory method allows for i) identifying the learning strategies that students follow in a FC, ii) examining differences in course performance among the student groups who applied different learning strategies, and iii) examining if those strategies change over time (course duration). The obtained results can help inform instructors on whether the FC design was effective in guiding students towards the set learning objectives. It can also help students increase their awareness of and allow them to reflect over their learning strategies. |
[CLOSING KEYNOTE] Professor George Siemens
For more information please contact stuart.dinmore@unisa.edu.au |
Is it possible to provide video links for all sessions please?
LikeLike
Only some of the sessions will have video links. these links will be sent via email to all those who register. thanks
LikeLike
so if I will be interstate and want to participate in the virtual classroom (where there is one on the schedule) I have to register now. Will there be video links other than the already listed virtual classrooms?
LikeLike
hi Howard. yes, you are right. you need to register for the session – then the virtual classroom link will be sent to all those who have registered.
The keynotes (and some selected sessions) will all be recorded and made available afterwards. Links will sent be out.
LikeLike
[…] following replay is from the University of South Australia Digital Learning Week [#dlw2016] with huge thanks to Shane Dawson and his team for hosting this fantastic event. All […]
LikeLiked by 1 person
[…] The following replay is from a talk at the UCL Institute of Education (Knowledge Lab) joint with UCL Interaction Centre. It is v2 of the talk, updating the one I posted earlier from University of South Australia Digital Learning Week. […]
LikeLiked by 1 person
[…] Digital Learning Week […]
LikeLike