Learning Designer[TM]: a theory-based SCORM-compliant content development tool.

This paper introduces a recently developed e-learning design and development tool called Learning Designer[TM] (version 1.0). Learning Designer assists learning designers and developers to build e-learning courses with Sharable Content Object Reference Model (SCORM) compliant learning objects using decision-making aids and recommended templates based on modern learning theories. Learning Designer assumes that learning is an active construction process as a learner participates in learning activities. Therefore, in this paper, a learning object is defined as a combination of content objects and learning activity objects. After six months of intense research on an e-learning design and development process, an explicit and standardized design and development process was identified, as was a tool to support the development process. Learning Designer includes four distinguishing features. First, it provides a step-by-step decision-making process for content design and development, with recommended templates based on learning theories and content design principles. Second, it employs customized learning sequencing of content and learning activity objects based on learning models. Third, it utilizes a guided metadata generator that meets the SCORM standards. Fourth, Learning Designer provides an embedded automatic manifest file generator that allows instructional designers to develop learning objects without having the technical knowledge to code manifest files for these learning objects.

**********

In recent years, many organizations in the public and private sectors have adopted e-learning solutions to replace traditional classroom training. Digital content, being produced using a learning-object oriented approach ported to e-learning courseware, has been very beneficial as a way to leverage the knowledge power of an organization. Utilization of the learning object concept in learning design and delivery can take the quality of learning to the next level of personalized learning. With these expectations, the digital content standardization movement has pursued accessibility, reusability, and interoperability of learning content since 1997. For several years, advancements in learning object concepts and standards have provided the potential and promise of learner-centered, anytime, anywhere learning opportunities for students. As more studies have been conducted, practical ways to resolve some of the problems and issues related to learning objects have emerged. However, most studies and related works of learning objects have focused on the technical issues rather than on learning issues. Therefore, the topic of learning issues, in the design of learning objects for a training/learning environment, is one of our study's objectives.

One major concern in meeting the performance needs of today's organizations is whether learning objects can satisfy not only low-level learning needs, such as information, knowledge, and skills acquisition through e-learning courses, but also high-level learning needs, such as performance support and just-in-time job aid solutions without formal course formats. One question arising from this concern is how to design and develop learning objects that are more flexible, reusable, and interoperable and still are highly effective in terms of learning aspects. The dilemma seems to lie between two extremes. At one extreme, it is necessary to satisfy technical requirements to be fully reusable, interoperable, accessible, extendible, and manageable; the opposite extreme is to satisfy the diverse learning needs of individual learners, instructors, tutors, and organizations. To resolve this difficult situation, the fundamental focus of utilizing learning objects should first be on organizational concerns, if such learning object initiatives are to solve performance issues within an organization (Tuso & Longmire, 2002). To be successful in the emerging e-learning space, for individuals as well as organizations, designers need to shift their thought processes away from the current, relatively static course-based way of thinking; they need to be able to design new content resources and reconfigure existing course content into dynamic digital forms. Designers and developers need to create media rich, personalized learning solutions that allow learners to take greater personal responsibility for managing their own information and learning needs while, at the same time, providing the relevant content to meet an organization's enterprise learning needs. One way of upgrading learning designers and developers would be to provide them with proper tools and retrain them to adjust to a new paradigm.

Learning design is a kind of apprenticeship in which skills and expertise are acquired only after understanding basic learning principles and techniques, assimilating general and domain-specific knowledge within a certain learning context, and inspecting good examples of teaching and learning practices. The field of learning systems design and development must share many aspects of science and art if quality educational programs are to be developed. Although most practitioners in the learning design field utilize learning theories and learning systems design models, there is always opportunity for design experts to apply their tacit and implicit knowledge and experiences to develop educational programs. During the course of the design and development process, skilled learning designers had some difficulties in articulating the kind of expertise they were using and how they were proceeding. In order for learners to master intended learning objectives and achieve an expected level of learning, the learning designers usually select and adapt certain theory-based learning models for educational programs. For novice learning designers, however, the task of program development is more problematic since they lack such experiences and expertise in learning design. In order to solve this problem, the researchers teamed up with two other organizations in Korea and built a tool to assist novice designers. The researchers thoroughly reviewed various learning theories as well as learning design models and identified 17 viable learning models and 37 learning activities. These models and activities are treated as Sharable Content Objects (SCOs) in the tool. Our goal in this paper is to present a theory-based Sharable Content Object Reference Model (SCORM)-compliant content development tool. Learning Designer is believed to improve not only the quality but also the quantity of sound e-learning content and programs--even when the designer does not know how to design a learning object-based program.

Learning Objects Standardization

The emergence of learning objects has changed many facets of e-learning. The objectives of e-learning and learning objects seem to match closely in that they both try to meet performance needs in a faster and more cost effective manner (Wagner, 2002). One remaining issue, however, is how to utilize learning object concepts to fulfill e-learning strategies at the organizational level as well as at the individual employee level. Current practices and standards for learning objects seem not to satisfactorily meet the needs of industry to combine possible learning solutions in order to achieve performance improvements. There are concerns about flexibility, ease of updates, searches, customization, and facilitation of competency-based learning when learning objects are adopted within a corporate learning environment (Longmire, 2002).

Although the concept of learning objects has not been fully standardized, technical standards to support a broad deployment of learning objects have been set by Advanced Distributed Learning (ADL, 2003). The SCORM is to provide a comprehensive suite of e-learning capabilities that enable interoperability, accessibility, and reusability of e-learning content. Since its first release in January 2000, there has been a dramatic evolution in the standardization of learning objects. This is due to the endeavors of many researchers and practitioners to clarify the incomplete concepts of technical standards of learning objects, such as definitions and terms of learning objects. Recently, ADL released the SCORM version 1.2 and version 1.3 application profile working drafts that specify the concepts, structure, and scope of learning objects in ways that are more practical. These recent updates include some new additions to expand the utility of learning objects. First, the most basic form of learning resources is defined. An asset is an electronic representation of media, such as text, images, sound, assessment objects, or any other piece of data that can be rendered by a Web client (ADL, 2003). A SCO is defined as a collection of one or more assets that can be launched and tracked by a learning management system (LMS) to support a learning objective. Sharable content asset (SCA) and SCO are independent of learning context so they could be reused in different learning objectives and experiences. Another addition, a content aggregation, which consists of activities and the intended sequencing of activities, is made by structuring activities in relation to one another and by associating sequencing information with each activity. Metadata represents a mapping and recommended usage for the assets, SCOs, activities, and content aggregations so they can be searched for and discovered within and across systems to further facilitate sharing and reuse. Currently, there are five places that metadata can be applied within a content package: manifest, organization, item, resource, and file. The current specifications of SCORM are not complete and need additional work on the definitions of the activities to accommodate various technical and content-related needs to use learning objects. The latest version, (1.3) seems to provide the minimum standards to develop instructional or learning modules that incorporate learning objects. It is expected that the newer versions may include more adaptable and applicable standards to develop, manage, and distribute learning objects to solve performance problems within organizations.

Need for a New Tool

As organizations move toward a reusable learning objects architecture for e-learning development and distribution, learning designers are being asked to suddenly change what they do and how they move from classical learning design to learning object-based design. They must acquire a different concept, process, and skill set if they are to migrate from being course and Web designers to learning object creators. For example, modularizing learning contents, metadata tagging, manifest file generation, and content aggregation are all new concepts. Therefore, there are hurdles learning designers must overcome to create learning objects. Meeting these new requirements will require that they adjust and upgrade their design and development skills in a relatively short period of time.

To solve this problem, three organizations in Korea--Ewha Womans University, a research organization, CyberMBA, a corporation and content developer, and 4Csoft, a software developer--have developed a new learning object-generating tool to support the seamless transition from the old learning design process to a learning object-based design process. After six months of thorough investigation of related literature and consultation with practitioners of e-learning design and development, an explicit learning design process has been identified and the support automation functions have been organized. As a result of our research, Learning Designer was developed and is now in the process of being field-tested. This e-learning content generating tool is intended to assist learning designers and developers to create SCORM compliant learning objects efficiently and effectively. It assists designers to enter metadata for all SCOs and assets using step-by-step guided templates. For content aggregation, the manifest generator is another feature of the tool to automate the process for those designers who do not know XML programming.

LEARNING DESIGN AND DEVELOPMENT PROCESS MODEL

Guidelines for Designing Learning with Learning Objects

In order to meet diverse needs and upgrade the current practices in learning and performance improvement using learning objects, the researchers explored effective and efficient ways to create these learning objects. SCORM-conformant reusable learning objects (RLOs) that offer greater efficiency and cost-effectiveness should follow the RAID criteria:

(1) Reusable: easily modified and used with different development tools;

(2) Accessible: can be searched and made available as needed by both learners and content developers;

(3) Interoperable: capable of operating across a wide variety of hardware, operating systems, and Web browsers; and

(4) Durable: do not require significant modifications for use with new versions of system software (Heims & Wagner, 2002).

Several learning design guidelines using learning objects were also used in this endeavor. First, as many researchers indicate (Baldwin & Ford, 1988; Rouiller & Goldstein, 1993; Holton, Bates, & Ruona, 2000) learning lacks value if application or transfer of learning does not exist during the course of the learning. Therefore, visible performance improvement is viewed as an ultimate goal of learning in using learning objects. Second, design criteria should include new standards that reflect the needs and perspectives of learners, instructors, and employers, and the learning context in which the learning takes place. Third, learning objects should contain information about the level of learning performance or expected outcomes because learning is always purposeful and intentional in its essence. Fourth, a learning process can be designed at a simple or more complex level according to the number of learning activities required to sequence the learning.

Content Structure and Delivery Mode

Content in instructional situations is more like information and is separated from learning. Therefore, content can have its own structures and delivery modes to enable better communication with learners. Creating a content structure using a mapping strategy is an essential first step. Learning Designer assists learning designers to create the content map either by content topics or by sessions. Another consideration is the type of content. Although it is hard to separate content from learning, in this paper content is categorized by facts, concepts, procedures, and principles according to Bloom, Gagne, Ausubel, and other researchers, as Figure 1 indicates (Reigeluth & Moore, 1999). Three different delivery modes--text & graphics, audio, and video-centered delivery--can also be selected for each content type depending on the quality of the infrastructures and the budget.

Learning Strategies and Sequencing

Since novice designers may have difficulty making a decision on a meaningful sequence of learning, Learning Designer might suggest two or three effective learning strategies with proper sequencing. With this support, novice designers can easily relate theories to learning strategies in the process of constructing learning objects. Seventeen learning models, such as virtual classroom, Student Teams-Achievement Division (STAD), action learning, etc., with sequenced activities are stored in the database, tagged by the variables shown in Table 2. These variables are used in a decision-making algorithm. In addition, each learning model includes recommended learning activities sequenced by the tool. The simplest sequencing might include the provision of learning content along with pre- and post-test, as needed. The next level, which focuses on performance, includes several learning activities such as presentation of content, note taking, reflection, and assessment activities. Possible learning activities at this level contain simulations, case studies, guided practice, problem solving, reflection notes, etc. This level includes more activities with the learned content and requires the learner's involvement in such activities. The third level entails cooperative and collaborative activities with peers and instructors. Group discussions, team-based decision-making, group case studies, and gaming are some of the representative learning activities. The fourth level involves more controlled and collaborative processes of learning to result in performance-oriented outcomes. Activities at this level include action-reaction, role-play, and project consulting. Conforming to the SCORM 1.3, the decision-making process for learning strategy with sequencing activities is an identical process to the content aggregation in objects-authoring.

Learning Design and Development Process

According to Wiley (2002), unless learning theory is incorporated into learning object construction and implementation, learning would neither be facilitated nor desired. Therefore, in the development of Learning Designer, a great deal of emphasis was placed on establishing the theoretical bases and learning principles of generating learning objects. As a result of our research into related learning theories, four stages that can be applied to the generation of SCORM objects (see Figure 3) were identified in the e-learning design and development process. The four stages illustrated in Figure 3 are rather open, flexible, and iterative. They should never be attempted as a linear or rigid process.

The development of content assets, which are the smallest components in SCORM objects, is explained first. In creating SCORM objects, designers and developers must generate content assets according to their blueprints of learning. In other words, multimedia content assets such as text, graphics, audio, and video are either newly produced or collected from a repository. The second stage is to create content objects by collecting and arranging the content assets. The learning content is categorized into fact, concept, procedure, and principle using Gagne, Merrill, and other content classifications (Bloom & Krathwohl, 1994; Merrill, 1983). Along with the type of content, the organizational needs or requirements, including budget, personnel, and resources, are gathered and used as a reference factor to select a delivery mode. Learning Designer provides templates for three different delivery modes--text and graphics, audio, and video. At this stage, designers may apply communication theories and interface design principles for content object creation. In order to assist this process, Learning Designer provides the designers with well-designed and theory-based templates according to the type of content and the delivery mode chosen. At the third stage, designers develop learning objects by sequencing content objects and learning activity objects to achieve an enabling objective. Enabling objectives are developed, through an authoring-aid function of the tool, to determine the type of learning content (well-structured vs. ill-structured), the level of outcome (knowledge vs. performance), and the level of learning difficulty (novice, intermediate, advanced). As learning designers complete these three stages, two or three recommended learning models, in addition to the related learning activities, are composed and suggested. Learning Designer contains 17 learning models in its database with appropriate learning activities to support each learning model. The 17 learning models are presentation, Web-based learning, memorization, learning through concept map, demonstration, storytelling, brainstorming, JIGSAW, STAD, simulation, role-playing, goal-based scenario, problem-based learning, inquiry-based learning, case-based learning, project-based learning, and action learning. At the same time, the 32 learning activities used for the 17 learning models, such as summary, discussion, brainstorming, reflection, evaluation, etc., are provided in the selection menu for designers to use as needed.

In Learning Designer, all content objects and learning activity objects are SCOs and can be edited using the developer's own editor and tagged with metadata. At this point, learning designers should determine learning strategies and models and select the learning methods, which will achieve the enabling objectives using a wizard. Finally, the course object to achieve terminal objectives is created by gathering several learning objects that had previously been developed. Terminal objectives are similar to enabling objectives in their characteristics. The only difference between them is the amount of learning. Terminal objectives comprise more learning in quantity and time than enabling objectives. Therefore, the process of creating course objects is very similar to that of creating learning objects in that it creates course objects by combining several learning objects. Learning Designer treats learning objects and course objects as SCOs. Therefore, metadata tags can be attached to those objects.

[FIGURE 3 OMITTED]

SYSTEM STRUCTURE AND FUNCTION OF LEARNING DESIGNER

System Structure of Learning Designer

Based on the design process model illustrated above, Learning Designer is constructed with five different databases to manage assets and objects (see Figure 4). Those databases are filled as learning designers progress in using the tool to create learning objects. Learning Designer contains two decision aids processes. One is used to create a content object. The second is used to create learning and course objects. When a learning designer or a content developer works on creating digital content, the system asks that he/she select a proper media type, such as basic text and graphics, audio, or video to meet the organization's needs. It also asks the designer to consider the type of the content to be delivered. The tool adopted Gagne's taxonomy--fact, concept, procedure, and principle--to classify the content types. According to the designer's decision criteria, the system suggests two or three recommended templates to be used. However, designers have freedom not to use those templates. For creating learning and course objects, Learning Designer consults with the designer by asking a series of questions such as the type of content (well-structured vs. ill-structured), level of outcome (knowledge vs. performance), and level of difficulty (novice, intermediate, advanced). After that, two or three recommended learning models that are arranged with proper learning sequences are provided. Learning Designer assists designers to tag metadata for SCOs. Content assets, content objects, learning activity objects, learning objects, and course objects are all categorized as SCOs, which can be tracked by an LMS in the SCORM run-time environment. The tool also provides an automated manifest file generator for the learning, course, and learning activity objects, which allows full interoperability of those objects with the SCORM run-time environment.

[FIGURE 4 OMITTED]

Functions of Learning Designer

To support learning design and development activities throughout the design process, several functions using web-based interfaces are provided. During the design and development process of a content asset, content object, learning object, and course object, Learning Designer gives users not only explicit assistance to design and develop learning content but also maximum flexibility and freedom to use the designer's creative and artistic ability. It assists users to enter metadata to all SCOs with step-by-step and template-based guides. For the manifest file generator, it automatically generates the XML code for those who do not know the programming language. The following illustrations are some sample screen shots that show the major functions and design processes of Learning Designer.

Content Assets and Content Object Generator

In the early stage of using Learning Designer, designers analyze needed content and build a content structure. In building this content structure, they need to categorize the content into hierarchies or chunks by the content classification scheme--fact, concept, procedure, and principle. After building the content structure, Learning Designer assists them to create each asset and gather these assets to make content objects. Each content object is packaged as an SCO with metadata.

[FIGURE 5A OMITTED]

[FIGURE 5B OMITTED]

Learning Object and Course Object Generator

In order to build a learning object, designers are asked to develop an enabling objective and decide the level of learning outcome such as knowledge or performance level. Other decision-making processes entail selecting the type of content structure (well-structured or ill-structured learning content) and the difficulty level of the enabling objective (novice, intermediate, or advanced level). The following figures are screen shots of the four steps of the decision-making process in building a learning object.

[FIGURE 6A OMITTED]

[FIGURE 6B OMITTED]

[FIGURE 6C OMITTED]

[FIGURE 6D OMITTED]

Once designers have completed the learning decision process up to this point, Learning Designer provides two or three recommended learning models with supporting learning activities to enable the designers to easily build a learning object. If a designer does not like any of these recommended learning models, he/she can choose another from the list of all 17 learning models. The designer also can customize his/her learning objects by combining any learning activity objects and content objects listed in the tool. After selecting or customizing a learning model, Learning Designer pulls up the proper learning activities supporting the selected learning model with recommended screen templates so the designer can build needed learning content from existing content repositories or directly create learning content through his/her own editors, which are capable of producing an SCO.

[FIGURE 7A OMITTED]

[FIGURE 7B OMITTED]

Since each content object and learning activity object is an SCO, it contains a metadata tag. All assets can have metadata to permit future reuse by other clients. The metadata generator minimizes the designer's workload by providing template-based guidance and explanations.

[FIGURE 8 OMITTED]

The next step of the learning object authoring process is packaging the learning object. As mentioned above, the learning object contains the sequenced content and learning activity objects as an e-learning module or course so it can communicate with any LMS using the SCORM content-to-LMS communications (CMI). In order for learning objects to be loaded into any SCORM-compliant LMS, the learning object must include an XML manifest file that shows the sequence of learning. Learning Designer generates this manifest file automatically using a default function. If a designer who is able to program XML wants to directly generate the manifest file, then he/she can edit this file manually using the source file. Course objects, the collection of several learning objects, can be generated in the same manner as creating learning objects. Learning Designer provides a process to develop a terminal objective for a course object. When a terminal objective is entered, designers can select existing learning objects to build a course object. The course object also contains a manifest file to be executed on a SCORM-compliant LMS.

[FIGURE 9 OMITTED]

Another feature of Learning Designer is its ability to add new learning models and activities, if needed. This means that the tool has been built upon an open architecture to maximize its functionality and flexibility. In order to enable the development of high quality content for multimedia learning, Learning Designer also provides a unique function to allow the editing of certain parts of learning content by media experts, such as graphic designers or multimedia specialists. For example, if a content developer finishes a learning object and needs to improve some graphic images inside that learning object, he/she can leave a note in a collaboration window asking what kind of graphics are needed and by what date. A graphics artist will be assigned to work on the request and post it back by the due date.

Values of Proposed Approach

Learning Designer is a promising tool to create highly personalized learning and just-in-time performance support. Although current specifications and standards of learning objects lack many features of meaningful learning concepts and models to completely satisfy the various learning needs of many clients, Learning Designer provides several advanced functions, reducing those gaps between the learning object standards and the various learning needs of the clients. First, the researchers' proposed approach to a learning object strategy provides meaningful connections and enhancements to existing standards for learning objects. The major contribution of the tool is the inclusion of the learning decision making process in developing learning objects that are compliant with the SCORM standards. In defining the content structure of a learning object, the researchers' approach proposes an extended interpretation of a learning object by adding a process to determine the levels of learning (knowing and performance) in the authoring of the learning objects. This is believed to help learners and instructors clarify their intent of learning at the very beginning stage when they are selecting or developing their learning objects. Once the level of learning is determined, the learners and instructors can move to the next decision process to select a learning model. Providing valid learning models and strategies during the decision making process is valuable because it allows more opportunities to apply and transfer learned content and competencies to each individual learner's jobs, tasks, and personal situations. Another benefit that Learning Designer provides is ease of development of the technical files needed for SCORM, such as metadata editing and manifest file generation. Without these kinds of extended functions, designers would experience a very acute learning curve because they would have to master XML, the programming language required to create the files required for the interoperability of the developed learning objects.

CONCLUSION

Learning Designer has been in usability testing at several organizations. In order to test the effectiveness and efficiency of this tool, quality e-learning courses developed by Learning Designer should be delivered and then evaluated. Several critical evaluation activities have been planned and implemented to enhance the tool. First, designers and developers who used Learning Designer were asked to write notes about its effectiveness as they used it. The actual design and development processes of some designers were systematically recorded using videotapes for qualitative evaluation data. A survey and interview is in plan to determine the performance and usability of the system. Several test cases using the tool are needed in order to analyze its strengths and weaknesses so that future improvements can be made.

The evolving concept of learning objects seems to make it possible to generate learning more customizable by assembling and reassembling learning objects. It also gives organizations a viable option to integrate training with knowledge management, human resources development, and enterprise resource management. It not only personalizes employee professional development, but also improves actual performance. With this promising future, the emerging role of learning designers will be in finding ways to create digital content that effectively and efficiently meets the requirements of RLO standards. Learning Designer is one such attempt to assist content designers and developers in easily generating sharable and reusable learning objects. It will be continuously modified and upgraded to meet client expectations for better and more user-friendly tools by incorporating effective and efficient design methods; this will improve the quality and quantity of both learning and e-learning products.

Bloom          Gagne         Ausubel     Anderson     Merrill

Knowledge      Verbal        Rote        Declarative  Remember
               Information   learning    knowledge    verbatim
Comprehension                Meaningful               Remember
                             learning                 paraphrased
Application    Intellectual              Procedural   Use a
               Skill                     knowledge    generality
Analysis       Cognitive                              Find a
Synthesis      strategy                               generality
Evaluation

Reigeluth           Learning
                    Designer[TM]

Memorize            Fact
information
Understand          Concept
relationship
Apply skills        Procedure
Apply generic       Principle
skills

Figure 1. Content Types Related to Instructional Taxonomies

Variables                  Values

Level of outcome           Knowing level vs. Performance level
Type of learning contents  Well-structured vs. Ill-structured
Focus of learning          Focused vs. Open
Grouping for learners      Individual vs. Group
Difficulty level           Novice, Intermediate, Advanced
Control of learning        Learner-centered vs. Instructor-centered

Figure 2. Variables for Selecting Learning Models

References

Advanced Distributed Learning. (2003). SCORM version 1.3, Retrieved April 15, 2003, from: http://www.adlnet.org/adldocs/Other/SCORMV1.3_AppProfile.zip.

Baldwin, T. T., & Ford, J. K. (1988). Transfer of training: A review and directions for future research. Personnel Psychology, 41, 63-105.

Bloom, B. S., & Krathwohl, D. R. (1994). Taxonomy of educational objectives. Handbook 1: Cognitive Domain. Reading, MA: Addison-Wesley.

Heims, F. & Wagner, E. D. (2002). Macromedia MX: Empowering Enterprise eLearning. Macromedia White Paper.

Holton, E. F. III, Bates, R., & Ruona, W. E. A. (2000). Development of a generalized learning transfer system inventory. Human Resource Development Quarterly, 11, 333-360.

Longmire, W. (2002). Emerging strategies for effective e-Learning solutions, Retrieved April 11, 2003, from: http://www.learnativity.com/download/LwoL3.pdf.

Merrill, M. D. (1983). Component display theory, In Reighluth, C. M. (Ed.), Instructional Design Theories and Models (pp. 279-333), Hillsdale, NJ: Lawrence Erlbaum.

Reigeluth, C. M. & Moore, J. (1999). Cognitive Education and the Cognitive Domain, In Reighluth, C. M. (Ed.), Instructional Design Theories and Models, V. II. (pp. 51-68), Hillsdale, NJ: Lawrence Erlbaum.

Rouiller, J. Z., & Goldstein, I. L. (1993). The relationship between organizational transfer climate and positive transfer of training. Human Resource Development Quarterly, 4, 377-390.

Tuso, G., & Longmire, W. (2002). Competency-based systems and the delivery of learning content, Retrieved April 11, 2003, from: http://www.learnativity.com/download/LwoL3.pdf.

Wagner, E. D. (2002, June 18). The new frontier of learning object design. The e-Learning Developers' Journal. Retrieved April 14, 2003, from: http://www.elearningguild.com/pdf/2/061802DST-H.pdf.

Wiley, D. A. (2002). Learning objects need instructional design theory. In Rossett, E. (Ed.), The ASTD e-Learning handbook, McGraw-Hill.

MYUNGHEE KANG

Ewha Womans University 

Korea

[e-mail]

DOO H. LIM

University of Tennessee 

USA

[e-mail]

MINKYUNG KIM

CyberMBA Corp.

Korea

[e-mail]