Generative learning objects (GLOs): design as the basis for reuse and repurposing Tom Boyle Director of the Learning Technology Research Institute (LTRI) London Metropolitan University Email:
[email protected]
Abstract There is considerable interest in the topic of reusable learning designs. These offer the prospect of capturing effective designs for learning and making them available for reuse and adaptation. Much of this work is focused at the level of lesson plans or above. However, there are many layers on which learning design works. Below the ‘lesson plan’ level these need to focus on learning activities for understanding key concepts and procedures. This paper, building on the work of the Centre for Excellence in Teaching and Learning (CETL) in Reusable Learning Objects, deals with reusable learning designs at this basic level. The RLO CETL has produced nearly 200 multimedia learning objects. It has recognised, however, that it needs to go beyond producing specific learning objects. The idea of capturing successful learning designs and making these the basis for reuse, rather than content, is at the core of the concept of generative learning objects (GLOs). The authoring and adaptation of generative learning objects is achieved through a specially developed authoring tool called GLO-Maker. This approach leads to improvements in productivity and the quality of the learning objects produced. Crucially, tutors can also use the tool to adapt existing GLO based learning objects to suit the local needs of their students. This paper will set out the need for GLOs, how these are developed using the GLOMaker tool, and the advantages of this approach over traditional approaches to learning objects. Finally, it will point to ongoing and future work both in enhancing the tool and linking this work as a service to other ‘higher’ layers of learning design.
Background: content and design-based approaches to learning objects The traditional approach to developing learning objects has focused on content, and standards for packaging and describing this content (e.g. IMS 2009, IEEE 2002, ADL SCORM 2009). Repositories of learning objects based on these standards are meant to improve learning:
2
“.. by making content more readily available, by reducing the cost and effort of producing quality content, and by allowing content to be more easily shared” (Duval et al. 2004)
This vision for learning objects has been to improve the quality of teaching and learning through the widespread availability of “self-contained” learning resources. Several national and international repositories have been developed. However, the evidence points to limited impact in achieving this aim (e.g. Koppi et al. 2004). There are several reasons why the learning object vision has failed so far to achieve its potential. However, one central reason is the failure in this approach to consider the central issue of the quality of learning. In the traditional, standards-oriented approach, there is really no guidance as to how to develop high-quality learning objects, either in terms of pedagogy or the design features that will facilitate reuse.
The traditional model for reuse has been to focus on content (IEEE 2002). However, content on its own is of very limited pedagogical value. The RLO-CETL has focused on a problem ignored in the main approach of standards-based learning objects – the design of high quality learning objects. Rather then assume that quality in learning would automatically arise out of availability for reuse Boyle (2008) argues that we need to tackle the central issue of the design and development of high quality learning objects in the first place:
“high quality design and development of learning objects is crucial before we get to issues of metadata and software packaging. The primary message … is good pedagogical design is at the heart of effective learning objects (Boyle 2008)”.
This approach and emphasis was adopted by the Centre for Teaching and Learning in Reusable Learning Objects (RLO-CETL). This five-year project, funded by the Higher Education funding Council for England focused from the beginning on high quality design (RLO-CETL 2009). In the initial phase of development (2005-2007) the RLO-CETL developed nearly 200 multimedia learning objects. These can be accessed from the CETL website (RLO-List 2009). The development of these learning objects used an ‘Agile approach’ (Boyle et al. 2006). The characteristics of this approach are that tutors (and often learners) work in small, empowered teams 3
with multimedia developers to create learning objects. These projects begin with the identification of real and significant educational problems. The projects were driven by the need to design high-quality resources that would help students to overcome these problems. Typically, an iterative prototyping development approach was used. The developers constructed initial versions of what they believed the tutors/learners wanted, which were then refined and improved until the final learning object was produced. These learning objects were normally then evaluated with large samples of students. This work has been extensively reported: see, for example, Boyle (2008).
There are three main problems with this approach:
The process of developing learning objects in this way is very intensive, and is difficult to scale-up;
The learning objects thus developed cannot be repurposed/adapted by tutors to meet their needs, without accessing the skills of a specialist multimedia developer;
Above all the primary emphasis in this approach is on design. However, the designs are embedded in specific learning objects, and were not available for independent reuse.
To tackle these problems, and produce a more powerful basis for reuse, the concept of generative learning objects (GLOs) was developed. With GLOs the primary focus of reuse is not the specific learning object but rather the pedagogical design that underpins the object. This switch in emphasis raises two main challenges. The first is the development of a clear conceptual model to capture and represent these designs. This issue is tackled in the following section. The second challenge is to make these designs accessible to tutors through a tool that permits the creation and adaptation of learning objects. A special tool, called GLO-Maker, has been developed to enable users to create and adapt generative learning objects. This is described in the central section of the paper. The paper then considers how GLOs fit into the wider framework of research into ‘design for learning’. The concept of layered learning design is used to relate this work to design at the ‘lesson plan’ level and above. Finally, the paper discusses current and future development work.
4
GLOs: a design-based approach to learning objects GLOs are concerned with the pedagogical designs underlying (successful) learning objects, and making these the primary focus for reuse. These underlying designs have to be rendered explicit in a conceptual model. These designs have to be rendered explicit in two distinct ways. The first form relates to human understanding. We need to articulate the often implicit decisions involved in design for learning. Furthermore, we need to render these decisions in a way that can be executed by computer software to produce learning objects based on the design. These problems are tackled using a form of representation borrowed and adapted from Generative Linguistics, in particular Systemic Grammar (Halliday 1973, 1975).
Generative linguistics is used as the basis for the GLO approach for two main reasons. First of all, it provide a generative form of representation, not a simply a descriptive form. Generative linguistic models seek to capture the processes involved in producing language (rather than simply describing the structures after they are produced). This is an important point. The description of an object after it has been finished can take a radically different form from a description of the processes involved in producing that object. In the GLO approach, we seek to capture the decisions made in producing learning objects in a way that can be inspected, adapted and reused by teachers and learners. Generative linguistics, furthermore, provides a basis for capturing these decisions in forms that are amenable to interpretation and execution by a computer. It thus, importantly, provides the basis for a form of representation that is both amenable to understanding by human users and executable by formal computer software.
There are many forms of generative grammar. It is not a matter of which, in some absolute sense, is the best. It is a question of choosing the representation that best meets our needs. The particular approach adopted for representing GLO structure is based on Systemic Grammar (Halliday 1973, 1975). Its approach of representing ‘deep structure’ as functional decisions, which are then mapped onto the forms of language, proved to be particularly productive in capturing and formalising decision processes involved in pedagogical design. This distinction between pedagogical function (what you want to achieve pedagogically), and the forms that realise that 5
function (how you achieve it), is central to the architecture of the GLOs. The paper will not provide a review of Systemic Grammar; rather it will show how these concepts have influenced the conceptual structures developed to represent GLOs.
The process of developing a GLO involves extracting and representing the design structure that underpins a series of learning objects. This is perhaps best illustrated by showing how the first GLO design pattern was developed. This was based on a set of learning objects that won a European Academic Software Award (EASA) in 2004. These learning objects may be accessed online at EASA (2004).
The first step in the process was to extract a series of screen layouts, or templates. A sequence of particular templates represented the ‘surface structure’ of the design. Many representations of design stop at this level, i.e. as a series of screen templates which can be ‘filled in’ with content added by the user. However, to develop a fuller representation the extraction process went further. It extracted and represented formally the pedagogical decisions underlying the generation of the learning object. This second level of abstraction does not have to do with form – it is rather concerned with pedagogical functions like ‘introduce’, ‘understand’, ‘test’ and how these are organised to represent a particular pedagogical design. These decisions are captured in a network that expands from left to right. The first level in the network represents the top-level decisions. Each node is then progressively refined, as far as required by the design. Figure 1 provides an example of one of these networks. This figure sets out the underlying decision structure for the EASA design. These networks are a very similar to the ‘design action potential’ (DAP) networks described in (Boyle 1997).
6
Figure 1: Network representing of the pedagogical design decisions underpinning the ‘EASA’ learning objects The first level represents the topmost, simplest division of the functions in the EASA pedagogical design. These are labelled as ‘Orient’ (the user to the learning task), ‘Understand’ (facilitate the learner understanding), and ‘Use’ (getting the user to apply their knowledge). At this level the EASA design is not very different from many others. Each of these high-level decisions, however, is further refined. The distinctive features of the EASA pattern are more obvious at the next level of refinement. In particular, ‘Understand’ is typically refined into getting the learner to ‘Apprehend’ a concept (gain an overall appreciation) before ‘Comprehending’ the components and relationships in the target domain. Finally, ‘Use’ is implemented normally as a scaffolded construction exercise (in the original EASA learning objects the learners assembled an example of the programming construct they had just learned). Each of the use functional decisions can be further refined as required.
The decisions are not represented as one rigid path, but rather as a network of choices. There is a main or default path through this network (represented by the highlighted choices in the network shown in Figure 1). However, it was obvious from reviewing the original learning objects that not all of the decisions were implemented in all of the learning objects. The decision tree represents a default path with permitted
7
variations available in this design pattern. Not all of the learning objects based on this design have to be structurally exactly the same. However, they all show key features in common.
The underlying structure represents a selected and organised set of pedagogical choices. This is captured as a decision structure – where each node represents a pedagogical function which can then be further refined/expanded as required.
These pedagogical decisions have to be realised in a form that can be communicated to a learner. This set of ‘deep structure’ pedagogical decisions is thus mapped onto a ‘surface structure’ which consists of a sequence of ‘page’ or screen layouts. Each of these page layouts organises and co-ordinates a series of component ‘containers’ (text boxes, media components). Content (e.g. text, pictures or animations) is then loaded into these component holders to complete the creation for a concrete learning object (e.g. See Figure 2 for examples of completed page layouts). The theoretical basis for generative learning objects (GLOs) is described in more detail in Boyle (2006).
Figure 2: Examples of the ‘surface’ structure ‘pages’ of GLOs as seen by learners. Clarifying the GLO conceptual architecture is only half the task. To achieve their potential the concepts need to be made accessible in an attractive and easy to use form. This problem was tackled by creating an authoring tool that would embed the GLO patterns, and make them available both for creating new learning objects and adapting existing learning objects.
8
GLO-Maker Authoring tool The GLO-Maker tool can be used to create specific learning objects based on a selected design. Each of these learning objects developed in this way can be repurposed by local tutors (or learners), using the same tool, to adapt the resources to their local needs and preferences. All the learning objects so created, or adapted, run as stand-alone Web based learning objects.
Version 1 of the authoring tool was released in July 2008. It is free to use and can be downloaded from the website (GLO-Maker 2009). The authoring tool provides two main interfaces. The first interface allows the user to access a pedagogical design expressed as a structured set of pedagogical choices. The second interface then expresses these choices as a sequence of screen layouts. These correspond broadly to manipulating the ‘deep structure’ functional choices and ‘surface structure’ realisation of these choices, as described in the previous section.
In the initial version of the authoring tool (GLO-Maker 0.5) the user could directly access a network representing the underlying functional choices. This is illustrated in Figure 3. The left hand panel represents the major choices made. The main panel then displays a small network of choices for refining the ‘Orient’ option. Making the choices explicit in this way means that they can be inspected, and the accuracy of the representation improved. Thus refinements in later versions meant that the distinction between ‘Quick’ and ‘Full’ orientation was removed, as it turned out to be redundant with the choices already available in the network.
In the In GLO-Maker version 1 this interface was changed to a drag and drop interface as illustrated in Figure 4. The main reason for this change was to make the interface more flexible and attractive to users. The underlying conceptual structure remains the same. However, its expression is modified in order to improve accessibility and ease of use.
9
Figure 3: Representation of functional choices in original version of GLO-Maker (version 0.5)
The user selects a pedagogical design from the drop down box on the menu bar ‘EASA’ has been selected in this screen shot’. The user can use this plan as it is, or modify it by dragging an option from the left hand panel, and inserting it at the chosen position in the sequence. This represents the top-level ‘storyboard’ of the design. Each design plan is represented by a core sequence plus a palette of choices which can be used to extend or modify that plan.
The tool is constructed to be extendable, so that as more designs are elicited from tutors they can be added to the ‘drop down’ box at the top. In the future, it is anticipated that designs will be inserted that are non sequential in structure.
10
Figure 4: Top level plan of the ‘EASA’ design in the GLO-Maker authoring tool When the top-level plan is completed the user can launch the second GLO maker interface. The plan automatically produces a sequence of page layouts that can be used to implement this ‘pedagogical plan’. This is illustrated in Figure 5a. The default page layout for a particular function can be overridden by the user selecting one of the alternatives illustrated in the bottom left hand panel. Crucially, attached to the page is functional advice on how to unfold this page in order to achieve the desired function. This is accessed through the ‘?’ on the title bar. The overlay screen shot in Figure 5a illustrates how advice is provided.
The user can then add text and upload media, including still pictures, Flash animations and videos, to create a concrete learning object. Figure 5b provides a screenshot from one such completed learning object as seen by the learner.
11
Figure 5a: Screen shots for realising the ‘Apprehend’ function in the EASA design pattern
Each learning object is saved in a separate package in its own directory. This contains all the resources needed for the learning object. The package consists of a ‘Player’, an XML file which contains the ‘script’ for the learning object produced, and a subdirectory which contains all the media assets used. The learning object can thus be played from any location into which this package is moved – for example a computer desktop, or a web server or embedded in a learning management system.
Crucially, all the learning objects developed in this way can be easily adapted. The learning object is simply loaded back into the authoring system. The learning object appears the same as in normal authoring mode. The user can then adapt the learning object either in small ways by changing text or graphics, or in larger ways by adding, deleting or modifying pages in the main sequence. This ability to modify the learning object is considered a crucial feature of the GLO approach. It allows local tutors to 12
adapt the learning object to meet the learning needs and preferences of their students. The modified learning object is simply saved out to its own named package, with the modifications captured in a new XML file.
Figure 5b: Completed screen from the EASA pattern as seen by the learner
This approach provides a very powerful way of capturing designs in a way that is accessible to tutors. However, the tool was constructed to be extendable, so that new designs elicited from tutors can be plugged into the tool. The next section outlines, through a case history, the addition of a new, exciting design.
Developing and evaluating new ‘plug-in’ designs The EASA learning design was extracted from a series of successful learning objects. The second example provides a more direct way of eliciting learning designs. The RLO-CETL runs a series of hands-on workshops were tutors, and often learners, work in groups to brainstorm and produce initial learning designs. The learning design discussed in this section originated in this manner. A group associated with the UK National Subject Centre for History, Classics and Archaeology produced an outline design at one of these workshops. The particular focus of this design was to get
13
across to students the idea that there may be multiple interpretations of the same artefact. Figure 6a shows the initial outline of the ideas together with a picture of the group working together.
The workshop was followed by collaboration between this group and members of the RLO-CETL to clarify and implement the GLO. This involved several rounds of discussion, clarification and refinement. Over the course of these meetings storyboards were drawn up and potential screen designs outlined and discussed.
Figure 6b shows a screenshot of one of the central screen layouts developed for this design. The main structure of the pattern is outlined in the panel on the left. The screen consists of three main components: there is a picture of the artefact to be discussed (in this case the Altar of Pergamon from the Acropolis). Above this there is a picture of various experts who comment on this design. To the right of the screen there are a series of topics which the learner can select to ask questions of the experts. The learner can thus choose a topic, ask a particular expert, and then compare with the views of another expert. It is up to the learner which questions they want to ask and in which order. This design worked well to meet the initial needs of the development group (see eMi GLO 2009). However, what makes it more interesting is that it is a generic design. Using the GLO-Maker tool, a teacher may select a topic of their own choosing, enter the questions they wish to be asked, and upload a series of pictures and audio files to represent the views of the experts. The pattern is independent of the particular content. In fact, some teachers have used this pattern to ask the students to create their own learning resource. The GLO maker website provides a tutorial on this learning design, together a zip file that contains all the resources necessary for recreating the learning object illustrated here (GLO-Maker 2009).
The link between the active elicitation of new designs from teachers and learners in workshops and the GLO-Maker tool is important. The tool has been constructed so that more designs can be added. This extensibility is an important feature of the tool. This is discussed further in the final section of the paper (on ongoing and future developments).
14
Figure 6a: Initial sketch of the ideas for the eMI design arising from an RLO-CETL workshop (picture from workshop inserted)
Figure 6b: Screen shot of the ‘Access Views’ page, from the full eMI design, as implemented in GLO-Maker version 1
15
The wider picture: the strategic implications of the GLO approach How do GLOs fit into the wider work on learning design and learning objects? At the broader level the concept of GLOs contributes in two strategic ways. Firstly, it posits a close relationship between learning designs and the learning objects. It views learning objects as generated from underlying designs. However, work in these two areas has largely proceeded in parallel. The work on GLOs suggests that they should be treated in a more integrated manner, in which learning objects are generated from, and represent particular instances of, learning designs. Secondly, it extends the focus on learning design below the ‘lesson plan’ level at which most of the work is focused. It thus raises the issues of different layers of learning design, and the relationship between these layers.
Learning objects and learning designs are two of the fundamental entities of the discipline of technology enhanced learning. Clarifying the relationship between these areas produces significant benefits for advancing the subject area. The approach in which learning objects are treated as instances of learning designs has been clarified in the rest of the paper. This section will thus focus on the relationship between GLO learning designs and designs which are focused on higher levels of teaching and learning, e.g. lesson plans.
The learning objects discussed in this paper focused on one clear learning goal or objective. Learning objects at this level are meant to represent basic units that can be combined and arranged to form higher order teaching and learning structures. The standard approach in learning object literature is to talk of ‘aggregation’. This approach is essentially descriptive: it does not clarify the relationship between these different layers of objects, other than to view higher order layers as combinations of more basic units (e.g. Verbert et al. 2005a Verbert 2008b). It also does not deal with the relationship between learning objects and the designs from which those objects are generated.
There is a parallel, distinct area of research that is focused specifically on ‘Learning Design’. This is a broad and vibrant area of work (e.g. Harper and Oliver 2002, Dalziel 2003, Britain 2004, Lockyer et al. 2008). Of particular relevance is the fact
16
that an international specification has been produced in terms of IMS LD (IMS-LD 2003). It is not the purpose of this paper to provide a comprehensive review of IMS LD. The purpose rather is to point to how IMS LD treats learning objects, and to propose an alternative model. IMS LD focuses on ‘learning designs’ primarily at the lesson plan level. Learning objects are treated as content which is loaded into these lesson plans. The GLO approach, however, suggests that significant learning design takes place at the basic learning object level. It thus immediately points to a separate layer of learning design. Far from learning objects simply being ‘content’ loaded into lesson plans, this perspective views design at the learning object level as crucial. The GLO project thus raises the important issue of different layers of learning design and how these layers relate to each other.
An initial representation of the conceptual space linking learning objects and learning designs is set out in Figure 7. This conceptual space is structured on two key dimensions. The first dimension, represented on the vertical axis, represents the dimension of size or scale. The second dimension, on the horizontal axis, represents the relationship between ‘object’ and ‘design’. These two dimensions, in turn, provide the framework for two key relationships.
In relation to the first dimension, that of size or scale, the key relationships is one of ‘service’. Each lower layer should provide a service to the layer above. What is more, there should not be confusion about the nature of the service provided. In IMS LD, for example, the ‘learning designs’ operate mainly at the session level. The relationship to the lower level is seen as one of loading learning object content. However, this short-circuits an important part of the conceptual space. Lesson plans, in themselves, are often too high-level to deal with specific learning problems, e.g. a student may struggle to master the basic concepts in mathematics. Learning object designs, or GLOs, provided potential solutions to these problems that may be incorporated into higher-order lesson plans. Teachers often do not have the time to think deeply about designing the solutions to all the learning problems of their students. This points to the need for sharing in communities of practice, where teachers may use the solutions developed by others, rather than trying to develop all the solutions themselves. Learning object designs or GLOs can be reused and adapted by tutors to fit the needs of their students. 17
Artefacts produced
Underlying designs Course plan
Actual course
Session plan
S E R V I C E
Actual Session
GLOs
Learning objects INSTANTIATON
Figure 7: Initial mapping of the conceptual space linking learning designs and learning objects
The second key relationship is one of instantiation – turning designs into specific learning objects or events. The GLO-Maker tool provides executable designs to make this easy at the learning object level. Furthermore, the tutor is not stuck with a particular fixed learning object. They can adapt the design and produce a new instantiation, a new adapted learning object, to meet their particular needs.
This conceptual model is introduced briefly here. There is not the space here to elucidate it in detail. However, that is not important at this stage. The important point is to emphasise that apparently disparate areas of work in learning objects and learning designs may be integrated in a unitary conceptual workspace. The exploration of the relationships in this workspace provides a basis for ongoing and future work. The present paper provides a detailed exploration of the instantiation
18
relationship at the learning object level (how objects are generated from learning designs). Further papers explore the second central issue of the nature of the layers and the relationship between them (Boyle 2009).
Ongoing and future work The ongoing and future work concentrates on three main strands. The first strand involves production of version 2 of the GLO-Maker authoring tool, due for release in July 2009. This is being developed in Adobe Flex, a powerful application development language (Adobe Flex 2009). GLO-Maker 2 is designed to have a ‘plug-in’ architecture. Developers may thus add their own components to expand the system. The new version of the tool will be open source, free for educational use, and available for download through the GLO-Maker website (GLO-Maker 2009).
The second strand links this technical development to ongoing creative work with tutors through workshops and other means. The aim is to incorporate more designs developed in these workshops into the tool. This link from the tool through to the active participation of teachers and learners introducing new patterns is considered crucial.
The third strand of work is more conceptual and theoretical. This concerns the topic introduced in the penultimate section of the paper – integrating work on GLOs into the broader research on learning designs and learning objects. The primary challenge is to produce a conceptual representation that provides an integrated framework linking learning object and learning design work. An initial focus of this work is to delineate different layers of learning design and to clarify the relationship between these layers. The primary focus in this work is to tackle conceptual integration at the level of pedagogy. This is in parallel, and complementary to, the work which focuses on technical standards. This work will be pursued partly through the major new LDSE (Learning Design Support Environment) project which aims to produce an integrated learning design support environment for teachers (LDSE 2009).
Acknowledgements: many people have contributed to the work described in this paper. I would like to acknowledge the core team who have worked on the
19
development of the GLO-Maker tool, especially Dejan Ljubojevic, Martin Agombar and Enzian Baur. This work has been carried out in the context of RLO-CETL. I would like to acknowledge the input of all my colleagues in the RLO-CETL especially Dawn Leeder from Cambridge University, who has actively supported and contributed to GLO developments from the beginning. The eMI design described in the paper was developed in conjunction with the UK Higher Education Academy Centre for History Classics and Archaeology. Eleanor O’Kell and Cary MacMahon, both originally based at the Centre, have played a central role in creating and developing this design.
References ADL SCORM (2009) http://www.adlnet.gov/Scorm/ Adobe Flex (2009) Adobe Flex website: http://www.adobe.com/products/flex/ AUTC Learning Designs (2002) Website for AUTC Learning Designs project http://www.learningdesigns.uow.edu.au/ Boyle T. (1997) Design for multimedia learning. Prentice Hall. Boyle, T. (2003) Design principles for authoring dynamic, reusable learning objects. Australian Journal of Educational Technology, 19(1), 46-58. Available online at: http://www.ascilite.org.au/ajet/ajet19/boyle.html Boyle T. (2006) The design and development of second generation learning objects. Invited talk at Ed Media. In E. Pearson & P. Bohman (Eds) Proceedings of Ed-Media 2006, World Conference on Educational Multimedia, Hypermedia & Telecommunications, June 26-30 2006; Orlando, Florida. Boyle T. (2008) The design of learning objects for pedagogical impact. In Lockyer, l., Bennett, S. Agostinho S. & Harper B. (Eds) The Handbook of Research on Learning Design and Learning Objects: issues, applications and technologies. Information Science Reference. Boyle T. (2009) Layered learning design. Paper accepted for CAL 2009. March 2009, Brighton, UK. Boyle T, Windle R., Leeder, D., Wharrad H., Alton, R & Cook J. (2006) An Agile method for developing learning objects. Proceedings of the 23rd annual ascilite conference: Who’s learning? Whose technology? Available online at: http://www.ascilite.org.au/conferences/sydney06/proceeding/pdf_papers/p64.pdf
20
Britain S. (2004) A Review of Learning Design: Concept, Specifications and Tools: A report for the JISC E-learning Pedagogy Programme, May 2004. Available at: http://www.jisc.ac.uk/uploaded_documents/ACF83C.doc Dalziel J. (2003) Implementing learning design: the learning Activity management system (LAMS). Proceedings of the 20th Annual Conference of the Australasian Society for Computers in Learning in Tertiary Education (ASCILITE). Available at: http://www.ascilite.org.au/conferences/adelaide03/docs/pdf/593.pdf Duval, E., Hodgins, W., Rehak, D. & Robson, R. (2004). Learning objects symposium special issue guest editorial. Journal of Educational Multimedia and Hypermedia. 13, 4, 331-342. AACE. EASA (2004). Learning Objects for Programming – winner of European Academic Software Award: http://www.londonmet.ac.uk/ltri/learningobjects/index.htm eMI GLO (2008) eMI GLO report at http://www.heacademy.ac.uk/hca/themes/elearning/emi_glo GLO-Maker (2009) Web site for the GLO-Maker authoring tool: http://www.glomaker.org Halliday M. A. K. (1973) Explorations in the functions of language. Edward Arnold. Halliday M. A. K. (1975) Learning how to mean. In Lenneberg E.H. (Ed.) Foundations of language development. UNESCO. Harper B. and Oliver R (2002) Information and Communication Technologies in Flexible Learning. Presentation of the outcomes of the AUTC project. Streamed audio presentation available at: http://www.iml.uts.edu.au/autc/ IEEE (2002) Draft Standard for Learning Object Metadata. http://ltsc.ieee.org/wg12/files/LOM_1484_12_1_v1_Final_Draft.pdf IMS (2009) IMS Global Learning Consortium Web site: http://www.imsglobal.org/ IMS LD (2003). IMS Learning Design specification V1.0. Available at: http://www.imsglobal.org/learningdesign/index.html
Koppi, T., Bogle, L. & Lavitt, N. (2004) Institutional Use of Learning Objects: Lessons Learned and Future Directions. Journal of Educational Multimedia and Hypermedia, 13, 4, 449-463. AACE. LDSE (2009) LDSE (Learning Design Support Environment) web site: http://www.tlrp.org/tel/ldse/
21
Lockyer, L., Bennett, S., Agostinho, S., & Harper, B. (Eds.) (2008) Handbook of Research on Learning Design and Learning Objects: Issues, Applications and Technologies. Information Science Reference.
RLO-CETL (2009) Web site of the Centre for Excellence in Teaching Learning (CETL) for Reusable Learning Objects: http://www.rlo-cetl.ac.uk/ RLO-List (2009) Online access to the learning objects developed by the RLO-CETL: http://www.rlo-cetl.ac.uk/joomla/index.php?option=com_content&task=view&id=246&Itemid=297
Verbert K., Jovanovic J., Gaševic D. & Duval E. (2005 a) Repurposing Learning Object Components, In R. Meersman, Z. Tari, P. Herrero (editors) On the Move to Meaningful Internet Systems 2005, Lecture Notes in Computer Science, Volume 3762 / 2005. Springer. Verbert K., Jovanovic, J., Gaševic, D., Duval, E., & Meire, M. (2005 b) Towards a global component architecture for learning objects: a slide presentation framework. In Kommers, P., & Richards, G. (Eds.), Proceedings of World Conference on Educational Multimedia, Hypermedia and Telecommunications 2005. AACE. All Web references checked on 30 January 2009.
22
