Printer Friendly

The Effects of Representation Tool (Visible-Annotation) Types to Support Knowledge Building in Computer-Supported Collaborative Learning.

Introduction

Collaborative learning is a method that enables learners to share each other's knowledge and foster collaborative knowledge construction (Eryilmaz, van der Pol, Ryan, Clark, & Mary, 2013). It is effective for enhancing the internalization of knowledge and promoting a higher quality of collaborative knowledge construction (Garrison & Arbaugh, 2007; Morgan, Whorton, & Gunsalus, 2000). In particular, learners use collaborative learning to share knowledge and solve problems in ill-structured problem-solving learning environments (Beers, Kirschner, Boshuizen, & Gijselaers, 2005). Communication activities that include professional knowledge sharing and different perspectives can enhance the level of learning performance (Lomi, Larsen, & Ginsberg, 1997). However, in the process of achieving higher-quality solutions, many restrictions can be imposed due to the learners' diverse perspectives in such areas as sharing problem awareness, negotiating a variety of opinions, and building collaborative knowledge based on communicative activities (Fuks, Pimantel, & Lucena, 2006). In addition, collaboration load can occur in computer-supported collaborative learning (CSCL) environments, and this can lead to ineffective learning processes and unsuccessful learning performance.

Eryilmaz et al. (2013) argued that collaboration load has consistently occurred in asynchronous discussions. This might have been due to the lack of a shared frame of reference to help team members identify which part of learning content was related to each opinion and individual claim in knowledge building (Hewitt, 2005; Suthers, Vatrapu, Medina, Joseph, & Dwyer, 2008). To build a shared frame of reference in support of online discussions, some annotation functions have been used. The annotation functions provided by "Knowledge forum" and "Concept map" have allowed learners to show all postings in chronological order at one time, write comments pertaining to the related learning content, and share their individual understanding of the learning content. However, when attempts have been made to refer to specific parts of the learning content, the shared frame of reference has not been sufficient to identify the diversity of perspectives and integrate the opinions of team members at one time, which could increase collaboration load.

To overcome these limitations, Eryilmaz et al. (2013) proposed linked annotation. This is an innovative artifact for online discussion in which there is an overarching notion of a challengeable function linking a participant's contributions with the entire related text provided by the other participants or an instructor. The linked annotation function can play a significant role in forming a shared frame of reference to draw higher-level solutions through meaningful communications. In particular, it effectively supports the building of a common ground and lowers the effort needed to identify the related learning content through minimized coordination. Some empirical studies exploring the use of linked annotation have shown that collaboration load decreases and interactions such as assertiveness and conflict increase during online discussions (Ding, 2009; Eryilmaz, Alrushiedat, Kasemvilas, Mary, & van der Pol, 2009; Muhlpfordt & Wessner, 2005). Other studies have found that linked annotation has no influence on promoting interactions aimed at clarification and interpretation, and that there is no meaningful effect on collaborative learning outcomes (Eryilmaz et al., 2013; van der Pol, Admiraal, & Simons, 2006). These diverse perspectives might have resulted from a failure to consider the collaborative knowledge-building process when designing a representation tool (Beers et al., 2005; Rummel & Spada, 2005; Slof, Erkens, & Kirschner, 2010; Suther & Hundhausen, 2003).

According to the collaborative knowledge-building process, knowledge-sharing activities that clarify and interpret the meaning of learning content are a prerequisite to knowledge- construction activities that negotiate disparate opinions and derive solutions (Beers et al., 2005; Zhu, 2012). Because the accuracy of shared knowledge has a significant role in building higher-quality constructed knowledge (Barron, 2003; Bromme, 2000; Cannon-Bowers & Salas, 2001; Clark & Brennan, 1991), the learning strategies for building shared knowledge and constructed knowledge should be differentiated. However, in most previous studies, learners have received the same learning methods during both the knowledge-sharing and knowledge-construction stages and the same communication activities have been implemented to build shared and constructed knowledge (Chuy, Zhang, Resendes, Scardamalia & Bereiter, 2011; Eryilmaz et al., 2013; Yucel & Usluel, 2016). Given that the sequential sharing of learning content from the key concept to more complex learning content can effectively enhance the accuracy of shared knowledge and lead to a higher level of constructed knowledge (Beers et al., 2005; Slof et al., 2010), more elaborate learning strategies that consider the collaborative knowledge-building process should be provided to build shared and construed knowledge.

This study was designed to investigate how representation tool types, designated "visible annotation," enhance the accuracy of shared knowledge and foster the level of constructed knowledge in CSCL environments. Based on the principle of collaborative knowledge construction, we developed the following three types of visible annotations with a linked annotation function: the TL-type, representing the two content-understanding learning phases necessary to build shared knowledge (TL); the TLL-type representing one concept-understanding learning phase and one content-understanding learning phase used to build shared knowledge (TLL); and the C-type, representing the one content-understanding learning phase used to build shared knowledge (C). The three visible annotation types include the same problem- solving learning phases used to build constructed knowledge.

Specifically, the following research questions were addressed in this study. First, what are the effects of the visible-annotation tool on the accuracy of shared knowledge in the CSCL environment? Second, what are the effects of the visible-annotation tool on the level of constructed knowledge in the CSCL environment?

Knowledge-building process of CSCL

Collaborative learning is effectively carried out if there are active interactions between team members sharing various opinions about learning content (Kolloffel, Eysing, & de Jong, 2011; Suthers & Hundausen, 2003). The process of sharing learning content and problem awareness from diverse perspectives in poorly structured problem-solving learning environments can help learners reduce the possibilities for misunderstanding and lead them to in-depth discussions (Beers et al., 2005; Garrison & Arbaugh, 2007). However, if the number of unshared messages among the learners grows, collaborative interaction is replaced by simple interaction, and this does not positively influence shared knowledge building and collaborative learning outcomes (Lomi et al., 1997; Suthers et al., 2008).

As Beers et al. (2005) and Rummel and Spada (2005) indicated, these limitations might be caused by a failure to consider the collaborative knowledge-building process when designing a representation tool. According to previous research, the limitations on representation tools may be affected if the collaborative knowledge building process is ignored (Slof et al., 2010; Suthers & Hundausen, 2003). Therefore, the appropriate learning processes and learning methods should be applied when carrying out the sequential sharing of learning content from the key concept to more complex learning content (Beers et al., 2005; Cannon-Bowers & Salas, 2001). In addition, the process of collaborative knowledge building should be considered when designing collaborative learning processes to stimulate effective knowledge-sharing activities (Barron, 2003).

Beers et al. (2005) introduced a model that explains the process of collaborative knowledge construction (see Figure 1). The model shows that knowledge shared between team members is constructed through a process of externalization and internalization. New knowledge is then constructed through negotiation and integration (Jorczak, 2011; Levesque, Wilson, & Wholey, 2001), which is important to the development of a new understanding based on the consolidation of shared knowledge. In contrast, Rummel and Spada (2005) proposed a collaborative learning model that enables learners to modify and elaborate their understandings through a repetitive process in which different individual perspectives are negotiated. This suggests that the processes of externalization and internalization must be carried out during the entire learning process, not simply before the shared knowledge stage (Ding, 2009; Gao, 2013).

It is therefore apparent that the stages of externalization and internalization are not necessarily prerequisites to the negotiation stage. Accordingly, externalization, internalization, and negotiation could conceivably be applied to each stage simultaneously throughout the entire learning process to build shared and constructed knowledge (Gao, 2013; Rummel & Spada, 2005). In addition, it could be argued that the process of identifying problem awareness and sharing individual understanding is required at the knowledge sharing stage, and furthermore that the negotiation process in which different individual opinions are shared to solve a problem is required at the knowledge construction stage (Beers et al., 2005; Jorczak, 2011; Levesque et al., 2001). The visible annotation developed in this study intends to provide a learning strategy that considers the collaborative knowledge-building process and includes the linked-annotation function to enhance the accuracy of shared knowledge and foster higher-quality knowledge construction for online discussion.

Difficulties of shared knowledge building

Although active interactions are indispensable to effective collaborative learning, it is not easy to derive accurate shared knowledge and higher-quality problem solving from these interactions (Eseryel, Ifenthaler, & Ge, 2013). Even when individuals are provided with the same learning tasks, the degree to which they understand the learning content can be very different. This can result in collaboration load when individuals share their understanding of the content and negotiate various options from different perspectives. Furthermore, these difficulties can expand in a CSCL environment as opposed to a face-to-face environment.

The difficulties of building shared knowledge in a CSCL environment can be explained as follows. First, collaborative learning requires coordinated activities for effective knowledge construction. Although the division of roles, technical support for performing the task, and determination of discussion time are not directly related to the task itself, they are crucial factors leading to fruitful collaborative learning (Gibson, 2001; Kirschner & Erkens, 2013). Therefore, supporting coordinated activities such as collaborative scripts or scaffolding can help learners diminish their collaboration load and engage in both knowledge sharing and construction.

Second, collaborative learning must consider the collaborative knowledge- building process to produce fruitful collaborative learning outcomes. The purpose of collaborative learning is to effectively carry out a complex task (Eryilmaz et al., 2013). However, learners generally experience difficulties with the complex task itself, and this can result in increased collaboration load (Kirschner, Paas, & Kirschner, 2009). In addition, the complexity of the learning content can lead to misunderstandings and result in inaccurate shared knowledge. Therefore, elaborate instructional designs for complex learning are necessary to reduce collaboration load and enhance the accuracy of shared knowledge, as the accuracy of shared knowledge can influence the level of constructed knowledge (Beers et al., 2005). As indicated by Levesque et al. (2001), meaningless interactions can negatively affect collaborative learning. Thus, carrying out shared knowledge-building activities through meaningful interactions based on the collaborative knowledge-building process is required.

Third, the representation tool should provide functions that are appropriate for the learning task (Kolloffel et al., 2011; Suther & Hundhausen, 2003). Given that the representation tool is the only means through which learners with various perspectives can express their opinions in CSCL environments, it should provide proper functions that consider the learning task if it is to lead to successful learning outcomes (Kirschner et al., 2009). A proper, shared frame of reference is essential to foster fruitful knowledge construction and greater gains in collaborative learning outcomes. Functions such as linked annotations and discussion threads support the building of a common ground and lower the effort needed to identify the related learning content. Therefore, it is necessary to design a representation tool that considers not only the collaborative learning process based on complex tasks, but also the distinctive functions that match the features of the learning tasks. The representation tool developed in this study is intended to focus on the collaborative knowledge-building process and the proper functions needed to carry out complex learning, thereby overcoming the difficulties experienced with shared knowledge building.

"Visible annotation" as a more viable representation tool

The development of computer technology has led to a recent interest in representation tools for externalizing individuals' knowledge within collaborative learning environments. A representation tool can help learners successfully reconstruct learning content and effectively deliver visualized information. In particular, representation functions are required at the troubleshooting stage in the CSCL environment because learners have difficulty sharing opinions and reaching fruitful solutions when they cannot efficiently track what is being discussed (Slof et al., 2010). The representation tool should therefore be supported in the process of learning problem solving skills by presenting annotations together with the related text, sharing opinions, and externalizing knowledge to further successful learning outcomes.

In this study, we developed "visible annotation" as a representation tool of text-based annotations using computer technology. The "linked annotation" function (Eryilmaz et al., 2013) was applied to connect annotations with their related learning content, leading to a shared, cognitive frame of reference and supporting higher quality argumentation. In particular, the tool was very effective in asynchronous discussions because tracking the annotations with their related learning content helped learners to accurately share various opinions and internalize knowledge learned from others. Moreover, the flow of the interactions could be clearly presented. Collaborative learning in problem-solving tasks included sharing learning content, negotiating various opinions, and arriving at solutions. Through this process, visualized information and troubleshooting strategies could be shared between team members and deeper knowledge construction was possible (Jorczak, 2011).

The proposed visible annotation was designed by focusing on the more elaborate learning strategies used to build shared knowledge and achieve higher solutions through collaborative learning. It consisted of a knowledge-sharing stage and a knowledge-construction stage. The TLL type of visual annotation used one concept-understanding learning phase and one content-understanding learning phase to build shared knowledge. This was devised to give learners an opportunity to sequentially share learning content from the key concepts to complex learning content. The TL type used two content-understanding learning phases to build shared knowledge. Learners achieved a basic understanding of the learning content in the first phase then deepened their understanding of it in the second phase. In the second phase, learners were instructed to focus on the difficulties that they had experienced in understanding the first phase. The C type used one content-understanding learning phase to build shared knowledge in the same way as the previous representation tools had done. The three types of visible annotations provided the same problem-solving learning phase to build constructed-knowledge. Tables 1 and 2 show the learning phases and learning methods based on the three different types of visible annotations.

Methods

Participants

Thirty-six students at a four-year college who had enrolled in one or more of four educational technology classes were invited to participate in this study. Sixteen of the students were women and 20 were men with an average age of 21.61. Sixteen were freshmen, 10 were sophomores, and 10 were seniors majoring in Educational Technology. The participants were randomly assigned to one of three groups, and each group was provided with different types of visible annotations for building shared knowledge in a CSCL environment; two content-understanding learning phases (TL); one concept-understanding learning phase and one content- understanding learning phase (TLL); and one content-understanding learning phase (C). The participants in each group performed the learning tasks in pairs. All of the groups carried out a "comprehension of learning content task" and a "lesson planning task" for four weeks within the CSCL environment.

Experimental materials

Pretests

Five multiple-choice problems were used to measure the participants' prior knowledge. A five-point Likert scale was developed to measure computer literacy and collaborative preference. Each test consisted of 10 questions (see Table 3). There were no significant differences between the groups in terms of their prior knowledge (F = 2.226, p = . 124), computer literacy (F = .660, p = .524), and collaborative preference (F = 2.580, p = .091).

Measurement

The independent variables consisted of the three types of visible annotation, and the dependent variables included the accuracy of shared knowledge and level of constructed knowledge. The effects of visible annotation for the enhancement of the shared knowledge were assessed using two computer- based tests, labeling and simple description, after completing the comprehension of learning content task for building shared knowledge. The tests consisted of five labeling problems and five simple pro and con descriptive problems based on previous research by Weinberger, Stegmann, Fischer, and Mandl (2007) (see Table 4). Pair- wise comparison was used and two evaluators were asked to rate the accuracy of shared knowledge on a scale ranging from "inaccurate" (0) in cases where both learners gave the wrong answers to "accurate" (1) in cases where both learners gave the correct answers. If one learner gave a partially correct answer and the other learner gave the correct answer, a partial score of 0.5 was given. Interrater reliability analysis revealed a Cronbach's alpha value of 0.92.

Constructed knowledge was analyzed by assessing the lesson-planning task for building constructed knowledge. Learners in each group were required to submit a lesson plan in pairs at the end of CSCL. A three-point Likert scale was developed to measure the level of the constructed knowledge, based on the research of Dick, Carey, and Carey (2004), and Gagne, Wager, Golas, and Keller (2004) (see Table 5). Three evaluators were asked to rate the level of constructed knowledge focusing on the instructional design principle on a scale ranging from "poor" (0) to "excellent" (1). If the lesson plan was rated "fair," a partial score of 0.5 was given. Interrater reliability analysis revealed a Cronbach's alpha value of 0.89. Table 6 shows the tasks and measures used to build shared and constructed knowledge.

Descriptions of visible annotation

The representation tool, visible annotation, was developed by focusing on previous studies of design principles used to develop representation tools and knowledge building theories (http://52.79.61.247). It was based on the Windows 2000 Server operating system. The MySQL Server 2000 was used as a database and the Web server used was CentOs 5.10 of the Windows 2000 Server. Python, HTML, CSS, and JavaScript were used as the programming languages.

The different types of visible annotations with the linked-annotation function consisted of different learning phases and learning methods based on the visible-annotation type used to build shared knowledge and the same learning phase and learning method used to build constructed knowledge (see Table 7).

The structure of the visible annotation consisted of a learning phase, learning content, annotation board, and FAQ. Figure 2 illustrates the menus and functions of visible annotation.

The concept-understanding learning phase focused on understanding the key concepts. If learners chose a key word from the learning content section, they could study concept-based knowledge by defining the meaning and explaining the pros and cons of key words. After submitting the concept-based learning task for a key word, annotations could then be presented on the annotation board and the key word in the learning content section could be linked automatically with the annotation. Therefore, if learners clicked on the key word in the learning content section, they could review its definition together with its pros and cons (see Figure 3).

The content-understanding learning phase focused on understanding sentence-based learning content. Learners could study the learning content in this phase by sharing various opinions through questions, explanations, and comments about it. If learners wanted to study specific aspects of the learning content, they could choose a sentence or a paragraph from it. After specifying the range of a sentence or paragraph, they could ask questions, explain what they understood, and provide comments about the specified parts. Annotations could be presented on the annotation board after submitting the content-based learning task. If learners clicked a specific sentence or paragraph from the learning content, it could be linked automatically with the annotation on the annotation board (see Figure 4).

The problem-solving learning phase focused on the lesson planning task at the knowledge-construction stage. Learners could share and negotiate various opinions through questions, explanations, and comments to derive solutions to complete the lesson-planning task. Other functions were provided in the same way as in the content-understanding learning phase (see Figure 5).

Procedure

The participants were divided into three groups, with each group being furnished with different types of visible annotations to perform their task. The experiments were conducted in the form of an assignment that lasted four weeks. The participants were first provided with a visible annotation of the CSCL environment and then participated in an experiment in pairs. An instructor provided the participants with special instructions for completing the task without having face-to-face contact.

Knowledge-sharing stage

All of the participants were provided with the same learning content using visible annotation in the CSCL environment. The learners carried out the comprehension of learning content task for shared-knowledge building. The learning content was presented as an online text related to learning theories, learning strategies, and instructional design principles. To understand the learning content, the TL group received two content-understanding learning phases. In pairs the learners studied the learning content by sharing their opinions through questions, explanations, and comments. The TLL group received two learning phases, one concept-understanding learning phase and one content-understanding learning phase. In the first phase, the paired learners studied the key concepts, defined the meanings and explained the pros and cons of the key terms in the learning content. In the second phase, they studied the learning content by sharing their opinions through questions, explanations, and comments about specific sentences or paragraphs in the learning content. The C group received one content-understanding learning phase. The learners studied the learning content in pairs, sharing various opinions. All of the learning tasks were conducted non- synchronously for two weeks. After completing the knowledge-sharing activities, each participant took a shared knowledge test on, for example, labeling and providing simple descriptions.

Knowledge-construction stage

The lesson-planning task for building constructed knowledge was conducted synchronously for two weeks. All of the learners engaged in the same problem solving learning phase and they shared and negotiated in pairs to achieve higher-quality solutions when carrying out the lesson-planning task. After the learners completed the knowledge-construction activities, the lesson plans were assessed by measuring the level of constructed knowledge (see Figure 6).

Results and discussion

The accuracy of the shared knowledge

We investigated the effects of using representation tool types to support knowledge building in the CSCL environment. The TLL group had the highest accuracy of shared knowledge and the C group had the lowest accuracy (see Table 8). The results of analysis of variance (ANOVA) of the shared knowledge found significant differences between the conditions [F (2, 15) = 5.543, p = .016] (see Table 9). A post doc Tukey test showed the difference between TLL and C [p = .015]. Although no significant differences were found between the other groups, there were differences between TLL and TL [p = .084] and between TL and C [p = .644].

Having accurate shared knowledge is very important because it can influence the quality of constructed knowledge (Cannon-Bowers & Salas, 2001; Dechurch & Mesmer-Magnus, 2010). According to previous studies, when learners are building accurate shared knowledge and integrating new understandings into their current knowledge base, inappropriate representations of the discourse and the complexity of finding their own comments can cause them to experience many difficulties within the CSCL environment (Hweitt, 2005; Simons, 2000; Suthers et al., 2008; Veerman, Andrissen, & Kanselaar, 1999). Visible annotation was intended to overcome these limitations by providing a shared frame of reference that could help learners enhance the accuracy of shared knowledge by indicating which part of the learning content was related to each annotation. Through the linked annotation function applied by visible annotation, visualized information and troubleshooting strategies could be shared between team members, and learners could identify the annotations and their related learning content. In addition, the linked annotation function facilitated the internalization and externalization of the learners' knowledge by representing the annotation with its related learning content.

TLL was found to contribute most effectively to enhancing the accuracy of shared knowledge. It was designed particularly to focus on building accurate shared knowledge, whereas most previous research has focused on enhancing the level of constructed knowledge in CSCL environments (Castek, Beach, Cotanch, & Scott, 2014; Chuy et al., 2011; Eryilmaz et al., 2013; Gao, 2013; Yucel & Usluel, 2016). Although many studies have found that the accuracy of shared knowledge plays a significant role in building higher-quality constructed knowledge (Barron, 2003; Bromme, 2000; Cannon-Bowers & Salas, 2001; Clark & Brennan, 1991), few efforts have explored the representation tools used to build the shared knowledge that leads to fruitful collaborative learning. TLL provided separate learning phases for concept and content understanding to build accurate shared knowledge based on a collaborative knowledge-building process. During the knowledge-sharing stage, learners could share different points of view on the key concepts and adjust misunderstandings related to learning content to establish a shared understanding. The results of this study showed that the learners' conflicting points of view declined throughout the sequential sharing process, leading to higher levels of shared knowledge. This resulted from their enhanced understanding of the meanings of things.

TL provided two learning phases for content-understanding through the process of asking, explaining, and commenting on learning content. The intent was for learners to share different points of view on sentence-based learning content in the first phase then deepen their understanding of the learning content in the second phase. However, the findings revealed that TLL was more effective than TL for enhancing the accuracy of shared knowledge. Previous research had shown that the sequential sharing of learning content from a key concept to more complex learning content could be effective for building accurate shared knowledge (Beers et al., 2005; Slof et al., 2010). Applying this to TLL, it was assumed that the first learning phase for shared knowledge building would influence the construction of an accurate concept understanding, and the second learning phase would influence the enhancement of deeper content understanding. Based on the results of this study, we suggest this assumption was correct and that TLL would be a suitable type of visible annotation for enhancing the accuracy of shared knowledge.

The level of the constructed knowledge

The TLL group had the highest level of constructed knowledge and the C group had the lowest level (see Table 8). The results of ANOVA of the constructed knowledge found significant differences between conditions [F(2, 15) = 9.838, p = .002] (see Table 9). A post doc Tukey test showed the difference between TLL and C [p = .001]. Although no significant differences were found between the other groups, there were differences between TLL and TL [p = .072] and between TL and C [p = 0.140]. The results of correlation analysis found significant differences between the shared and constructed knowledge [p = .021]. The results implied that the accuracy of the shared knowledge could affect the level of the constructed knowledge (see Table 10).

In this study, TLL was found to be the most effective for enhancing the accuracy of the shared knowledge and the level of the constructed knowledge. These findings were consistent with previous research and provided empirical evidence on how the quality of constructed knowledge is affected by the accuracy of shared knowledge (Barron, 2003; Bromme, 2000). Knowledge-sharing activities that clarify the meaning of learning content are a prerequisite to knowledge-construction activities that negotiate various opinions and derive solutions to achieve a high level of constructed knowledge (Levesque et al., 2001). The results of this study demonstrated that the quality of the collaborative knowledge was affected by the accuracy of the shared knowledge. The findings extend the previous finding that at the knowledge- sharing stage, the sequential sharing of learning content from a basic concept to complex content (considering the collaborative knowledge building process) enhanced the accuracy of the shared knowledge and resulted in a higher quality of constructed knowledge.

The correlation between shared and constructed knowledge was examined on the basis of quantitative correlations and according to collaborative knowledge construction theory (Beers et al., 2005). Significant correlations were found between the shared and constructed knowledge, implying that the accuracy of the shared knowledge could affect the level of the constructed knowledge. These results supported the findings of previous research on how the accuracy of shared knowledge leads to a higher quality of constructed knowledge. The results of this study have overarching implications for instructional design. They suggest that visible annotation with a concept-understanding learning phase used to facilitate knowledge building in the CSCL environment is significantly correlated with building an accurate shared mental model to understand learning content. In other words, providing a concept-understanding of a learning phase can promote understanding of the task itself, which builds accurate shared knowledge and higher-quality knowledge construction.

Conclusions and limitations

The representation tool can help learners with various perspectives share their opinions with others, build shared knowledge, and construct collaborative knowledge in CSCL environments. However, despite these advantages, the conventional representation tool has limitations. It cannot help shared knowledge reach a higher-quality cognitive domain because the processes leading to problem awareness, opinion sharing, and collaborative troubleshooting have not been fully considered in the conventional representation tools. In this study, visible annotation, which had the function of linking learners' annotations with related learning content, was developed to overcome these limitations and enhance the accuracy of shared knowledge and the level of constructed knowledge based on collaborative knowledge construction theory.

We proved that TLL with one concept-understanding learning phase and one content-understanding learning phase during the knowledge-sharing stage was the most effective tool for building shared and constructed knowledge and ensuring fruitful collaborative learning. This study demonstrated that visible annotation with a concept-understanding learning phase for shared knowledge building could positively enhance the accuracy of shared knowledge because the concept-understanding activities could promote comprehension of the key concepts and learners could deepen their content understanding during content- understanding activities. In addition, these findings suggested that the accuracy of shared knowledge could lead to higher-quality constructed knowledge. These findings are consistent with previous studies, which have suggested that the accuracy of shared knowledge is an important factor in enhancing the level of constructed knowledge in the CSCL environment (Beers et al., 2005; Cannon-Bowers & Salas, 2001; Eryilmaz et al., 2013; Levesque et al., 2001; Slof et al., 2010). Based on these findings, it can be concluded that TLL with learning activities, in accordance with each learning phase and based on building knowledge theory, can facilitate the accuracy of shared knowledge and the level of constructed knowledge in CSCL environments.

This study might have been limited by its structure. First, 36 students participated in this study, not a small number. However, the number of participants had to be greater to generalize the findings for further study. Second, the study measured the correlation between shared and constructed knowledge to identify the process of collaborative knowledge construction, effectively acting on the visible annotation tool developed for this study. Future studies should explore the processes of building shared and constructed knowledge in each group to achieve successful collaborative learning outcomes.

Acknowledgements

This work was supported by a National Research Foundation of Korea grant funded by the Korean Government [NRF-2014S1A3A2044609].

References

Barron, B. (2003). When smart groups fail. The Journal of the Learning Sciences, 12(3), 307-359.

Beers, P. J., Kirschner, P. A., Boshuizen, H. P. A., & Gijselaers, W. R. (2005). Coercing knowledge construction in collaborative learning environments. In Proceedings of the International Conference on CSCL 2005 (pp. 8-17). Taipei, Taiwan: International Society of the Learning Sciences.

Bromme, R. (2000). Beyond one's own perspective: The Psychology of cognitive interdisciplinari In P. Weingart & N. Stehr (Eds.), Practicing Interdisciplinarity (pp. 115-133). Toronto, Canada: Toronto University Press.

Cannon-Bowers, J. A., & Salas, E. (2001). Reflections on shared cognition. Journal of Organizational Behavior, 22(2), 195-202.

Castek , J. , Beach , R. , Cotanch , H. , & Scott , J. (2014). Examining middle- school students' uses of Diigo annotations to engage in collaborative argumentative writing. In R. Anderson & C. Mims (Eds.), Digital tools for writing instruction in K-12 settings: Student perception and experience. Hershey, PA: IGI Global.

Chuy, M., Zhang, J., Resendes, M., Scardamalia, M., & Bereiter, C. (2011). Does contributing to a knowledge building dialogue lead to individual advancement of knowledge? In H. Spada, G. Stahl, N. Miyake, & N. Law (Eds.), Connecting computer-supported collaborative learning to policy and practice: CSCL2011 conference proceedings. volume i--long papers (pp. 57-63). Hong Kong: International Society of the Learning Sciences.

Clark, H. H., & Brennan, S. E. (1991). Grounding in communication. Perspectives on socially shared cognition, 13(1991), 127-149.

DeChurch, L. A., & Mesmer-Magnus, J. R. (2010). Measuring shared team mental models: A Meta-analysis. Group Dynamics: Theory, Research, and Practice, 14(1), 1-14.

Dick, W., Carey, L., & Carey, J. (2004). Systematic design of instruction. Upper Saddle River, NJ: Pearson

Education.Ding, N. (2009). Visualizing the sequential process of knowledge elaboration in computer-supported collaborative problem solving. Computers & Education, 52(2), 509-519.

Eryilmaz, E., Alrushiedat, N., Kasemvilas, S., Mary, J., & van der Pol, J. (2009). The Effect of anchoring online discussion on collaboration and cognitive load. In Proceedings of AMCIS 2009 (pp. 576). San Francisco, CA: Americas Conference on Information Systems

Eryilmaz, E., van der Pol, J., Ryan, T., Clark, P. M., & Mary, J. (2013). Enhancing student knowledge acquisition from online learning conversations. International Journal of Computer-Supported Collaborative Learning, 8(1), 113-144.

Eseryel, D., Ifenthaler, D., & Ge, X. (2013). Validation study of a method for assessing complex ill-structured problem solving by using causal representations. Educational Technology Research and Development, 61(3), 443-463.

Fuks, H., Pimentel, M., & De Lucena, C. J. P. (2006). RU-Typing-2-Me? Evolving a chat tool to increase understanding in learning activities. International Journal of Computer-Supported Collaborative Learning, 1(1), 117-142.

Gagne, R. M., Wager, W. W., Golas, K. C., & Keller, J. M. (2004). Principles of instructional design (5th ed.). Belmont, CA: Wadsworth Publishing.

Gao, F. (2013). A Case study of using a social annotation tool to support collaboratively learning. The Internet and Higher Education, 17, 76-83.

Garrison, D. R., & Arbaugh, J. B. (2007). Researching the community of inquiry framework: Revise, issues and future directions. The Internet and Higher Education, 10(3), 157-172.

Gibson, C. B. (2001). From knowledge accumulation to accommodation: Cycles of collective cognition in work groups. Journal of Organizational Behavior, 22(2), 121-134.

Hewitt, J. (2005). Toward an understanding of how threads die in asynchronous computer conferences. The Journal of the Learning Sciences, 14(4), 567-589.

Jorczak, R. L. (2011). An Information processing perspective on divergence and convergence in collaborative learning. International Journal of Computer-Supported Collaborative Learning, 6(2), 207- 221.

Kirschner, P. A., & Erkens, G. (2013). Toward a framework for CSCL research. Educational Psychologist, 48(1), 1-8.

Kirschner, F., Paas, F., & Kirschner, P. A. (2009). A Cognitive-load approach to collaborative learning: United brains for complex tasks. Educational Psychology Review, 21(1), 31-42.

Kolloffel, B., Eysing, T. H., & de Jong, T. (2011). Comparing the effects of representational tools in collaborative and individual inquiry learning. International Journal of Computer-Supported Collaborative Learning, 6(2), 223-251.

Levesque, L. L., Wilson, J. M., & Wholey, D. R. (2001). Cognitive divergence and shared mental models in software development project teams. Journal of Organizational Behavior, 22(2), 135-144.

Lomi, A., Larsen, E. R., & Ginsberg, A. (1997). Adaptive learning in organizations: A System-dynamics-based exploration. Journal of Management, 23(4), 561-582.

Morgan, R., Whorton, J., & Gunsalus, C. (2000). A Comparison of short-term and long-term retention: Lecture combined with discussion versus cooperative learning. Journal of Instructional Psychology, 27(1), 53-58.

Muhlpfordt, M., & Wessner, M. (2005). Explicit referencing in chat supports collaborative learning. In T. Kosschmann, D. Suthers, & T. W. Chan (Eds.), Computer supported collaborative learning: The next 10 years (pp. 460-469). Mahwah, NJ: Lawrance Erlbaum.

Rummel, N., & Spada, H. (2005). Learning to collaborate: An Instructional approach to promoting collaborative problem solving in computer-mediated settings. Journal of Learning Sciences, 14(2), 201- 241.

Simons, P. R. J. (2000). Towards a constructivistic theory of self-directed learning. Self-Learning, 1-12.

Slof, B., Erkens, G., & Kirschner, P. A. (2011).Coordinating collaborative problem-solving processes by providing part-task congruent representations. In Proceedings of the International Society of the Learning Sciences (pp. 675-682). Chicago, IL: International Society of the Learning Sciences.

Suthers, D. D., & Hundhausen, C. D. (2003). An Experimental study of the effects of representational guidance on collaborative learning processes. The Journal of the Learning Sciences, 12(2), 183-218.

Suthers, D. D., Vatrapu, R., Medina, R., Joseph, S., & Dwyer, N. (2008). Beyond threaded discussion: Representational guidance in asynchronous collaborative learning environments. Computers & Education, 50(4), 1103-1127.

van der Pol, J., Admiraal, W., & Simons, P. R. J. (2006). The Affordance of anchored discussion for the collaborative processing of academic texts. International Journal of Computer-Supported Collaborative Learning, 1(3), 339-357.

Veerman, A. L., Andriessen, J. E. B., & Kanselaar, G. (1999). Collaborative learning through computer-mediated argumentation. International Society of the Learning Sciences, 31(4), 77-86.

Weinberger, A., Stegmann, K., Fischer, F., & Mandl, H. (2007). Scripting argumentative knowledge construction in computer-supported learning environment. In F. Fischer, I. Kollar, & J. M. Haake (Eds.), Scripting computer-supported collaborative learning (pp. 192-211). New York, NY: Springer.

Yucel, U. A., & Usluel, Y. K. (2016). Knowledge building and the quantity, content and quality of the interaction and participation of students in an online collaborative learning environment. Computers & Education, 97, 31-48.

Zhu, C. (2012). Student satisfaction, performance, and knowledge construction in online collaborative learning. Educational Technology & Society, 15(1), 127-136.

Yoonhee Shin, Dongsik Kim * and Jaewon Jung *

Department of Educational Technology, Hanyang University, Seoul, South Korea // shinyoonhee06@gmail.com // kimdsik@hanyang.ac.kr // jjungj5@gmail.com

* Corresponding authors

(Submitted August 23, 2016; Revised February 1, 2017; Accepted February 4, 2017)

Caption: Figure 1. Process of collaborative knowledge construction (Beers et al., 2005, p. 9)

Caption: Figure 2. Structure of visible annotation (TLL)

Caption: Figure 3. Concept-understanding learning phase

Caption: Figure 4. Content-understanding learning phase

Caption: Figure 5. Problem-solving learning phase

Caption: Figure 6. Learning processes in the CSCL environment
Table 1. Learning phases based on visible annotation types

  Visible                   Knowledge-sharing stage
annotation
    type             1st week                   2nd week

TL             Content-understanding      Content-understanding
               learning phase for         learning phase for deep
               basic understanding of     understanding of learning
               learning content           content

TLL            Concept-understanding      Content-understanding
               learning phase for         learning phase for
               understanding of key       understanding of learning
               concepts                   content

C              Content-understanding learning phase for
               understanding of learning content

  Visible        Knowledge-construction stage
annotation
    type             3rd and 4th week

TL

TLL            Problem-solving learning phase
               for completing the lesson-plan

C

Table 2. Learning methods for each learning phase

Type of learning phase     Learning method

Concept-understanding      To define the meaning and explain
                           the pros & cons of key terms

Content-understanding      To ask, explain, and comment on the
                           sentence-based learning content

Problem-solving            To negotiate various opinions and
                           derive solutions for completing the
                           lesson-planning task

Table 3. Sample questions for pretests

Type of pretest        Questions                # of items

Prior knowledge        Choose the wrong              5
                       explanation about
                       the brainstorming.

Computer literacy      I have no                    10
                       difficulties in
                       learning using the
                       computer.

Collaborative          I prefer to exchange         10
preference             opinions with my
                       colleagues.

Table 4. Sample questions for shared knowledge-test items

Category        Questions               # of items

Labeling        An expedition to             5
                study something at
                first hand

Description     To explain the pros          5
                and cons of
                "problem-based
                learning"

Table 5. Sample indicators to measure constructed knowledge

Category     Questions                # of items

Frame        The learning                  8
             objective is clearly
             stated.

Content      Lesson plan offers            5
             learning activities
             appropriate to the
             learning strategy.

Table 6. Tasks and measures to build shared and constructed
knowledge

Stage                 Task

Knowledge-sharing     Comprehension of
stage                 learning content task

Knowledge-            Lesson-planning task
construction stage

Stage                 Measure

Knowledge-sharing     Shared knowledge test
stage                 Two computer-based tests; labeling
                      and simple description after
                      completing comprehension of the
                      learning content task

Knowledge-            Constructed knowledge test
construction stage    Lesson-planning task based on
                      instructional design principles

Stage                 Reference

Knowledge-sharing     Weinberger et al. (2007)
stage

Knowledge-            Dick et al. (2004)
construction stage    Gagne et al. (2004)

Table 7. Learning phases and learning methods based on visible
annotation types

                           Knowledge-sharing stage
  Visible          Comprehension of learning content task
annotation
   type               1st week                      2nd week

    TL       Content-understanding         Content-understanding
             learning phase                learning phase
             To ask, explain and           To ask, explain and
             comment on the sentence-      comment on the sentence-
             based learning content        based learning content

    TLL      Concept-understanding         Content-understanding
             learning phase                learning phase
             To define the meaning         To ask, explain and
             and explain the pros &        comment on the sentence-
             cons of key terms             based learning content

     C       Content-understanding learning phase
             To ask, explain and comment on the sentence-based
             learning content

                Knowledge-construction stage
  Visible           Lesson planning task
annotation
   type               3rd and 4th week

    TL       Problem-solving learning phase
             To negotiate various opinions
             and derive solutions for
             completing the lesson-planning
             task

    TLL

     C

Table 8. Mean and standard deviation of the shared knowledge and the
constructed knowledge

               Accuracy of shared   Level of constructed
Visible            knowledge            knowledge
annotation
type             M         SD           M          SD

TL              7.42       .80        8.58        2.99
TLL             9.17       .75        11.25       1.22
C               6.50      2.23        6.33        .75

Note. N = 12. M: Mean. SD: Standard deviation. TLL: Two learning
phases for concept-understanding activities and
content-understanding activities during the knowledge-sharing
stage. TL: Two learning phases for content- understanding
activities during the knowledge-sharing stage. C: One learning
phase for content-understanding activities only during the
knowledge-sharing stage.

Table 9. ANOVA of the shared and constructed knowledge

Type of knowledge     Source           SS     df     MS       F

Shared knowledge      Group-inter    22.69     2    11.35   5.543
                      Group-in       30.71    15    2.05
                      Total          53.40    17

Constructed           Group-inter    72.69     2    36.35   9.838
knowledge             Group-in       55.42    15    3.69
                      Total          128.11   17

Type of knowledge     Source           p

Shared knowledge      Group-inter    0.016
                      Group-in
                      Total

Constructed           Group-inter    0.002
knowledge             Group-in
                      Total

Note. N = 12. SS: Sum of squares. MS: Mean square.

Table 10. Correlation between the shared and constructed knowledge

Type of knowledge     Source                    Shared
                                               knowledge

Shared knowledge      Pearson correlation          1
                      coefficient
                      p

Constructed           Pearson correlation         .537
knowledge             coefficient
                      p                           .021

Type of knowledge     Source                   Constructed
                                                 knowledge

Shared knowledge      Pearson correlation          .537
                      coefficient
                      p                            .021

Constructed           Pearson correlation            1
knowledge             coefficient
                      p

Note. N = 12.
COPYRIGHT 2018 International Forum of Educational Technology & Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Shin, Yoonhee; Kim, Dongsik; Jung, Jaewon
Publication:Educational Technology & Society
Article Type:Report
Date:Apr 1, 2018
Words:6867
Previous Article:Teaching Classical Mechanics Concepts using Visuo-haptic Simulators.
Next Article:Modelling and Simulating Electronics Knowledge: Conceptual Understanding and Learning through Active Agency.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters