**2. Theoretical background**

The learning process has been studied by learning psychologists in behaviorism, cognitivism, or constructivism; however, learning mainly occurs in our brain which is basically non-

<sup>1</sup> Since there is a kind of e-learning that is for self-learning based on e-contents with little to no interpersonal interaction, this research specifies interaction-based e-learning to focus more on highly interactive e-learning. Computer and network technology have contributed into two ways in teaching and learning; one is for developing effective contents, and the other is for enhancing interpersonal interaction. Contents-based self-learning, pretty popular in Korea, is based on oneway e-contents mostly in the form of VOD (video-on-demand) or WBI (web-based instruction). To eliminate the contents factor, the course developed in this research was interaction-based, with the minimum pre-developed contents.

observable. Only the consequences of learning can be observed. So many studies have dealt with the learning consequences rather than the learning process. However, in many cases, those studies generally turn out to be statistically insignificant when one tries to be rigorous about the learning outcome [6]. Therefore, [5] suggests that we should use the term, knowl‐ edge-building rather than learning, especially in regard to collaborative and interactive learning. According to Stahl [4, 5], knowledge-building is more tangible, concrete, and descriptive than learning. This term, knowledge-building, seems to include the whole process of external activity influencing on learning as well as internal learning itself within the brain. With care and practice, the knowledge-building process can be observed directly and empir‐ ically, because it accounts for externally observable activities and artifacts as experiential evidence. Therefore, we will use the term knowledge-building instead of learning to specify the observable and empirical approach to this research.

the interaction or strategies are effective has not yet been studied enough. There are some conceptual and theoretical articles on knowledge-building [3, 4]. However, there is a lack of theoretical research on a knowledge-building process based on empirical implementation. To find out more effective instructional strategies in interactive distance learning, we need to first understand how the knowledge-building process works in interaction-based e-learning1

Knowledge-building is said to be differentiated from *learning* [3, 5]. In [3], "Learning is an internal, unobservable process that results in changes of belief, attitude, or skill. Knowledgebuilding, by contrast, results in the creation or modification of public knowledge"(p. 1371). [3] also described that "knowledge-building environments enable ideas to get out into the world and onto a path of continual improvement in a form that allows them to be discussed, interconnected, revised, and superseded"(p. 1372). It focuses more on building knowledge-inthe-world as opposed to knowledge-in-the-head. To understand how the process mechanism is going and to find out better instructional strategy in interaction-based e-learning, the observable knowledge-building process rather than the internal learning process would be more useful. Also, the knowledge-building process should be more clearly disclosed in an interaction-based learning environment with active social communication, rather than

The purpose of this chapter, therefore, is to articulate a knowledge-building process in interaction-based e-learning. This research is concerned with how individuals and groups build their knowledge and construct meaning in interaction-based e-learning. This research also considers learning outputs, such as cognitive achievement or satisfaction levels, for better

This research will focus on a process-oriented approach; such an approach is focused on '*where* it makes a difference' rather than '*whether* it makes a difference'. Many other studies that utilize a learning *product-oriented* approach—such as comparative studies of learning results with the application of certain strategies—showed various differences on their effec‐ tiveness. However, it is said to be no significant difference by meta-analysis of each research result [6]. This 'no significant difference phenomena' indicates that research needs to shift from finding differences to reasoning the cause of differences. This research focuses more on *process* than product, and presents a more meaningful contribution in the theory and practice of

The learning process has been studied by learning psychologists in behaviorism, cognitivism, or constructivism; however, learning mainly occurs in our brain which is basically non-

1 Since there is a kind of e-learning that is for self-learning based on e-contents with little to no interpersonal interaction, this research specifies interaction-based e-learning to focus more on highly interactive e-learning. Computer and network technology have contributed into two ways in teaching and learning; one is for developing effective contents, and the other is for enhancing interpersonal interaction. Contents-based self-learning, pretty popular in Korea, is based on oneway e-contents mostly in the form of VOD (video-on-demand) or WBI (web-based instruction). To eliminate the contents factor, the course developed in this research was interaction-based, with the minimum pre-developed contents.

resource-based instruction for basically individual learning.

150 E-Learning - Instructional Design, Organizational Strategy and Management

interactive e-learning.

**2. Theoretical background**

comprehensive understanding of the knowledge-building process.

.

One who tried to disclose the knowledge-building process in interactive e-learning, such as Computer-supported Collaborative Learning (CSCL), was Stahl [4]. Stahl presented a diagram of a knowledge-building process in CSCL from theoretical discussion. His diagram consisted of two circles: one of personal understanding and the other of social knowledge-building (see Figure 1). Stahl [4] described the diagram as "*The convention in the diagram is that arrows represent transformative processes and that rectangles represent the products of these processes: forms of knowledge. To take this limited representation too seriously would be to reify a complex and fluid development—to put it into boxes and to assume that it always follows the same path. In particular, the diagram gives the impression of a sequential process whereas the relations among the elements can take infinitely varied and complex forms. Indeed the identification of the particular set of elements is arbitrary and incomplete. Perhaps despite such limitations and potential distortions the diagram can provide a starting point for discussing a cognitive theory of computer support for knowledge-building. It remains to be seen if such a phase model provides the most useful representation (In [4], pp.71)*."

He explicitly considered the relationship of processes associated with individual minds to those processes considered to be socio-cultural. The significance of Stahl's model is that he indicates the importance of social learning, which is considered to be essentially different from individual self-learning. He suggests that knowledge would be shared and constructed by social interaction in a CSCL environment. He is taking a social constructivist's perspective in which his work impresses upon a sequential process to knowledge-building and provides a starting point for discussing cognitive theory of CSCL as indicated in his research [4]. However, his model was derived from theoretical discussion and it wasn't verified by empirical evidence. As he mentioned in several papers [4, 5, 7], the research community should elaborate upon the knowledge-building process model by utilizing empirical research.

Besides, many studies on modeling the learning process have been reported [8]. Most of them, however, use a face-to-face learning environment or do not utilize empirical evidence. Moreover, they present linear learning procedures and do not consider other factors such as influential relationship between process and product. Therefore, a study on a comprehensive knowledge-building process that considers the learning output variables, such as cognitive achievement and satisfaction levels, as a form of empirical evidence is needed.

**Figure 1.** A diagram of the knowledge-building processes [4].

### **3. Method**

In order to articulate a knowledge-building process, an interaction-based e-learning course at college level was developed. Four experts in instructional design practice verified the course program throughout the development process. The subject of the program was 'General Understanding of Distance Education'. Fifty-six juniors at K Cyber University in Korea, from ages 20 to 50, were required to partake in team interactions assigned in the class for a period of eight weeks. Throughout the course, the students' perceived knowledge-building process was collected and analyzed from questionnaires. Cognitive achievement tests, satisfaction queries, online messages, interviews, and participatory observation data were also collected and analyzed. Various methods of statistical analysis including correlation analyses, contents analyses, t-tests, frequency analyses, etc. were applied to the data as well. For students' perceived knowledge-building process, each student was required to describe his/her own knowledge-building process four times during the course. Four times investigations helped how students' perception changed throughout the course progress. Average return rate of the each questionnaire was 84% out of all 56 students. Question items were as follows:

**•** How many hours per week do you spend for this course?


**Figure 1.** A diagram of the knowledge-building processes [4].

152 E-Learning - Instructional Design, Organizational Strategy and Management

In order to articulate a knowledge-building process, an interaction-based e-learning course at college level was developed. Four experts in instructional design practice verified the course program throughout the development process. The subject of the program was 'General Understanding of Distance Education'. Fifty-six juniors at K Cyber University in Korea, from ages 20 to 50, were required to partake in team interactions assigned in the class for a period of eight weeks. Throughout the course, the students' perceived knowledge-building process was collected and analyzed from questionnaires. Cognitive achievement tests, satisfaction queries, online messages, interviews, and participatory observation data were also collected and analyzed. Various methods of statistical analysis including correlation analyses, contents analyses, t-tests, frequency analyses, etc. were applied to the data as well. For students' perceived knowledge-building process, each student was required to describe his/her own knowledge-building process four times during the course. Four times investigations helped how students' perception changed throughout the course progress. Average return rate of the

each questionnaire was 84% out of all 56 students. Question items were as follows:

**•** How many hours per week do you spend for this course?

**3. Method**


For evaluation reliability, three evaluators graded 10% of the students' answer sheets and their responses were correlated (Pearson r=0.84, p<0.01). In terms of satisfaction level, a satisfaction measurement tool that was developed and validated by [9] was used after modification (reliability alpha=0.93).

After the course, 10 students were interviewed by telephone for 1–2 hours to verify all the quantitative data and to provide more detailed information regarding the factors of the knowledge-building process. The interview first started with the questions similar to the questionnaires in a flexible and unstructured manner and proceeded naturally to verify the knowledge-building models (an observable action model and a conceptual model reflecting the action model and other previous literature). All interview data were recorded and analyzed afterwards.

In the end, the conceptual model was verified by five experts (Ph.Ds. in Education) and ten students in the class. Respondents used a 5-point Likert scale (5 = fully verified, 1 = not verified), which was developed based on previous literature [1, 10] to rate validity, explicability, usability, generality, and comprehensibility of the conceptual model. Average rate of experts was 4.30 and average rate of students was 4.11 out of 5.00.

This study followed Rubinstein's [11] proposed procedure for modeling the knowledgebuilding process. According to Rubinstein, the modeling procedure is to achieve a simple high level of abstraction. So the procedure needs to be iterative until we get an abstractive pattern. In considerations of Rubinstein's perspective, the procedure in this research was as follows; Development of a treatment instructional program → Implementation of the program and collecting data → Coding questionnaire data → Deriving a rough pattern of the knowledgebuilding process → Analysis of learning output variables → Correlation of learning output variables and each stage of the knowledge-building process → Verification of the knowledgebuilding process action model by interview with learners → Conceptualization of knowledgebuilding process and its actual visualization → Verification of the conceptualized knowledgebuilding process model by experts and learners → Production of the verified conceptual diagram of the knowledge-building process in interaction-based e-learning. The more detailed procedure is shown in Table 1.


**Table 1.** Research procedure of this study.
