Multimodal Data for Learning
I am on the review board for this special call for papers on "Multimodal Data for Learning" for the Journal of Computer Assisted Learning (JCAL). The special issue deals with new data sources coming from the Internet of Things (IoT), wearables, eye-trackers and other camera systems, self-programmable microcomputers such as Raspberry Pi and Arduino. How can these multimodal datasets that combine traditional learning data with different data from physical activity, physiological responses or contextual information be exploited for learning?
On these pages you find information about my personal and professional background as well as some features about my interests in technology-enhanced learning, knowledge creation, knowledge transfer, and networked universities.
Learning technology has come a long way, and provides organisations, learners, and teachers with enormous opportunities to innovate not only their technical environment, but also the teaching and learning methodologies as well as their business processes. Most of all, the introduction of new media technologies leads to reflections about inherited traditional systems versus new approaches. My work contributes to these reflections and pursues not only innovation but also the effects that this innovation has on education, learning, and people.
This is a fast moving area and research focus changes quickly and often unexpectedly. My publications page contains a list of works in the field over the years. To a great extent they too reflect the changing nature of education.
My views on technology-enhanced learning
I am a passionate believer in the opportunities that technology has to offer to the knowledge society, both in terms of enhancement of learning and in reaching out to new learners. In remote and rural communities it is often the only way for people to access higher education. However, I am also of the opinion that technology alone does not produce new knowledge or learning and that new developments need to have a pedagogic and learner-centred approach.
Experience in commerce and education has shown that online solutions are at their best when built upon a traditional well-established structure. The pedagogic concept of Blended Learning is increasingly supported by universities and governments who realise that it provides a more sustainable approach than purely online offerings. This goes some way towards recognising that we cannot ignore pedagogic concepts that have been successful for decades before the internet arrived and still are.
Areas of interest
Technology enhanced learning has made giant leaps forward over the past few years. In my work, I try to keep up-to-date with latest developments and newest technologies. My current research interests focus especially on Learning and Knowledge Analytics, Language Technologies for learning support, Learning Networks, and Mobile Learning, but I also have a keen interest in other topics, including open education, game mechanics, or the most recent debate about connectivism.
Read more about my views on e-learning developments.
PLEs and Personal Services
Student-centredness has moved into the next phase with the emergence of personal learning environments (PLEs). In plain English, it’s a shift from centralised to de-centralised service provision. An interesting and more extreme viewpoint was expressed by some experts claiming that universities should not fulfil the role of an ISP (Internet Service Provider). This could lead to the outsourcing of many IT services to external providers and students. Especially with mobile technologies we already see an increased expectation by institutions that students have their own devices and should use them also for their studies.
Another aspect is that of a personal learning space, presenting students as active participants in control of their own tools - largely Web 2.0 community tools such as del.icio.us, weblogs, wikis, concept maps and search engines. To some extent this view represents the maturing information society, where knowledge is up for grabs and knowledge production takes a shortcut to publication.
From the institutions’ perspective, it was the personal teaching environment that up till now held the monopoly in service provision. The assumption had been that what benefits the tutor will cascade to benefit the learner. Of course, this led to a dependency of students on their sometimes technically challenged tutors. In a reverse case scenario a digital divide between students is the result, where a perfectly computer literate learner cannot use and develop their skills because the tutor resists change.
Especially with the emergence of Learning Networks and self-directed learning in communities, being able to determine your own personal environment is a step into the right direction.
Re-usable Learning Objects
Re-usable Learning Objects
What are Learning Objects? Lecturers' hand-outs, draft minutes of committees, telephone contacts, files, research papers? This question has been the topic of much debate.
The Learning Object model is based on the idea of creating independent interchangeable educational content which the user can sequence in different ways to produce individualised learning. The connection with object-oriented programming (OOP) has long been criticised and rejected. A number of attempts have been made to define Learning Objects, however, no general agreement has been reached.
Taking from different approaches, the following characteristics for Learning Objects in my opinion are the essential markers. They...
- need to be digital: This is for practical reasons only to distinguish from e.g. museum exhibits.
- have to have metadata (e.g. for discovery and contextualisation)
- should connect to learning outcomes: that is they need to be more than just files and would contain e.g. instructional information
- need to be potentially reusable and durable (see below)
This is the area of most debate. Opinions to this end vary from whole courses down to file level (or even code level). Every institution tends to apply their own preferences.
For reasons of clarity and pragmatism, I adopted a relatively widespread understanding of granularity by educationalists, which looks at RLOs as the smallest cognitive unit with a learning objective.
An RLO's value comes from its reusability in as many contexts as it can be appropriately applied. The value can be identified as:
value = number of potential contexts * appropriateness
Appropriateness here means the suitability to the learning context in terms of level of study, discipline, currency, methodology, or otherwise. The formula allows us to evaluate RLOs.
The dilemma of reusability of RLOs comes from the inherent context description. To be reusable in as many contexts as possible the RLO has to be context neutral and pedagogically neutral, so it can be customised by tutors to other pedagogic settings. At the same time it needs to contain appropriate pedagogic context descriptions within which to utilise the RLO.
This apparent contradiction can be partially overcome by appropriate cataloguing, which also needs to take into account the relevant metadata "standards" (LOM, SCORM, IMS, etc.). These are rapidly developing and requests for more metadata are continuously added. Merging schemes are also likely to appear, e.g. between RLOs and learning activities. There is some danger that RLOs would have to be re-catalogued continuously as standards develop.
When considering flexible sequencing of RLOs which would allow students to determine their own path through the learning materials, it's been realised that some of the most basic techniques and habits no longer work with RLOs. For example, a lecturer would normally refer to other lessons and make connections to ideas. Illustrative examples would be carried over across a number of learning units. Furthermore there are difficulties with review exercises. This would not work as the tutor would not know which units the student has followed.
When student sequencing and development of context-free RLOs are taken into account, the importance of instructional design becomes apparent. Rules and LD standards need to be built into the creation process to assure RLOs fulfil the expectations of re-usability. There is a strong training requirement for content creators (lecturers) implied in this. Authoring practices are likely to have to undergo changes. The impact on the content creator is difficult to determine but a critical risk factor for the success.
A Learning Object repository like a library only becomes sustainable when sufficient use can be anticipated. To build up useful reusable stock - even if all the human resources required are in place (editorial, cataloging, IPR, instructional design, etc. etc.) will take time. A change in the culture of authoring and an awareness campaign are probably needed to harvest the benefits.
Sosteric & Hesemeier (2002): When is a learning object not an object; IRRODL
Sicilia & Garcia (2003): On the concept of usability and reusability of learning objects; IRRODL
Three Pillars of e-Learning
e-Learning, or technology-enhanced learning (TEL) as it is called today, is a key concept for today's competitive higher education institutions. In fact, without e-learning some networked and virtual universities would not exist. Traditional campus-based universities as well as open and distance learning institutions such as the Open University had to redefine themselves to make room for technologies in their operations and provisions. Institutions are on the way of transforming their traditional roles and that of their members of staff and students. My involvement in state-of-the-art research into technology-enhanced learning, knowledge management, knowledge creation and transfer aims at pushing the envelop even further and innovating our approaches to education.
Although there is no unanimous definition of e-learning, the impact it has on Higher Education institutions can be divided into the following three areas:
To make e-learning an integral part of an institution it has to be firmly embedded within its business processes and strategic developments. This requires an enormous cultural change that not only affects lecturers as is often believed, but goes right from the top management to all sectors of the institution. It is a transformational change affecting everyone.
Cultural change on this scale takes time and effort. The usual picture is the bell curve with the fast adopters at the forefront of development, followed by the mainstream and tailed by those reluctant or unable to change. Mainstreaming a development always comes with a training requirement. A good strategy incorporates these training and migration issues.
This refers to the technology and the connectivity it provides. Proper e-Learning architecture is built on interoperability between different systems, for example an institution's student record system and a virtual learning environment. In the UK the e-learning architecture is underlying what is called a Managed Learning Environment or MLE. As the name suggests, MLEs allow institutions to better manage their (e-)learning through interconnected systems.
To make e-learning systems work together, interoperability standards have taken a prominent place in e-learning architecture developments. They play an important part in the connectivity of systems but also in future proofing the institutional MLE. Beyond the home institution, these emerging standards aim to be an important facilitator for inter-institutional collaboration.
A lot of attention is given to re-usable learning objects or RLOs. These are chunks of learning that can be easily shuffled around to create new paths of learning without re-inventing the wheel over and over again.
It is a fact that technology alone does not create learning. Because of the fast pace with which new tools are emerging, institutions are struggling to keep on top in their commitment to be pedagogically led not technology driven. In reality, though, effective use of a tool comes long after the tool has been created.
Blended learning has become a popular concept since more institutions want to combine their traditional teaching with an element of e-learning. Very few organisations today follow the online only trend that dominated the first enthusiastic approaches. This parallels the experience in e-commerce.
Currently, e-learning is mostly treated separately in terms of quality assurance, assessment, professional development and other areas. However, I firmly believe that at some stage it has to become integrated with generic teaching, in the same way as we do not treat the use of an overhead projector in class as a different approach. The important aspect is how learning is achieved, not which tools you use.
Quality assurance in e-learning is an important part for the success in this area. Online students need to have a comparable experience to those sitting in the classroom. It has become clearer that solely putting content online does not work for most students if there is no guidance and support. The MIT therefore had no business difficulties in offering all their learning material for free through the OCW initiative.
Quality of online delivery is closely connected to staff development in the area of online pedagogy. Special skills and awareness of the limitations of the technology are an important part in conducting successful e-learning courses. A rather new trend looks at learning activities and the mapping of these in an online environment. I discuss this in more detail together with learning objects.
Re-usable Learning Objects in e-Learning are mainly servicing the creation and sequencing of content artifacts for single self-directed learners. In HE, learning arises from interacting with peers and tutor via multi-learner activities. Although lesson planning is an integrated part of delivery, this still is largely a blind spot in e-learning.
In simple words the relationship between Learning Objects and Learning Activities is one of content and what to do with it, i.e. its instructional application. Suggestions have been made to identify the additional metadata needed for the re-use of Learning Activities and IMS LD is a specification to address this purpose.
What is important to take note of is that teaching strategies can be easier transferred to different educational settings than content. If designed efficiently, content objects can be pulled into a learning activity thus making it applicable in various contexts.
What's more, commercial studies have shown that recycling concepts not content is more cost efficient. Products like the Australian LAMS™ platform and the recognition of the wider possibilities for reuse of Learning Activities over Learning Objects cannot be ignored.
It should be noted that it is possible to build a learning activity sequence without any content (discursive and reflective in nature or a template sequence).
Dalziel (2003): Discussion paper for Learning Activities and Metadata
Feldstein (2003): How to design recyclable learning objects