Citekey: @Oshima2006

Oshima, J., Oshima, R., Murayama, I., Inagaki, S., Takenaka, M., Yamamoto, T., … Nakayama, H. (2006). Knowledge-building activity structures in Japanese elementary science pedagogy. International Journal of Computer-Supported Collaborative Learning, 1(2), 229–246. doi:10.1007/s11412-006-8995-8


Efforts to introduce KB into Japanese elementary classrooms: created two practical design principles—idea-centered lesson and collective knowledge advancement—based on 12 KB principles. In two design studies, these two practical principles were integrated with Japanese science activity structures. Considerations of local contexts when introducing KB is a sensible choice. In two design studies, they evaluated enactments of KB principles based on student participation. Two strategies reflected in this study: 1) it is challenging to target all KB principles at once; some are naturally more challenging (e.g., rise-above) but achieving one principles would open the door to another one; 2) treating teaching as a “design science” (REF) is an important approach for KB. In this study, recognizing study 1 demonstrated little awareness of community knowledge, the researchers adapted the activity structure in study 2 and added elements of offline classroom discussions. At the end of this study, they recognized a new challenge of coordinating scientific inquiry (e.g., build an experiment) and building conceptual knowledge. These two strategies are valuable.


The purposes of our study are: (1) to improve Japanese elementary science curriculum using knowledge-building practices, and (2) to contribute to the advancement of development principles for designing knowledge-building communities in classrooms. (p. 2)

First, we describe common Japanese elementary science activities and how they differ from knowledge-building practices (Scardamalia, 2002). Second, we discuss our redesign of Japanese elementary science lessons as knowledge-building practices by modifying and coordinating elementary science activities with a Computer-Supported Collaborative Learning (CSCL) technology called Knowledge Forum. Finally, we report two design studies of modified elementary science lessons. (p. 2)

Japanese elementary science activity structures: An established culture of learning (p. 2)

Lessons in Japanese schools have activity structures that are established through repeated research lessons (Rohlen & LeTendre, 1995). (p. 2)

Linn, Lewis, Tsuchida, and Songer (2000) videotaped and analyzed ten science lessons in five elementary schools in the Tokyo region. From their analysis, they found eight typical activity structures. (p. 2)

Depending on their students’ characteristics and classroom circumstances, science teachers in Japanese classrooms plan their lessons by using these activity structures. Each activity structure could function to facilitate the creation of a community of learners (Brown & Campione, 1996). When we as Bdeep constructivists’’ (Scardamalia & Bereiter, 2002) sit in the classroom, however, we rarely see students engage in that kind of knowledge advancement. (p. 3)

CoL vs KB
interesting mentioning that these activity structures could help CoL but not KB. (p. 3)

The first reason is that the activity structures identified by Linn et al. (2000) are not necessarily coordinated with each other to consolidate the classroom as a community of learners. (p. 3)

As Brown and Campione (1996) point out, many failures in structuring the classroom as a community of learners stem from the fact that instructional designers do not have a systematic view on how to create a community of learners in classrooms. (p. 3)

The second reason that these activity structures are not creating knowledgebuilding communities in Japanese classrooms is that not all communities of learners are necessarily knowledge-building communities (Scardamalia, 2002; Scardamalia & Bereiter, 2002). (p. 3)

to facilitate different types of learning. In a community of learners, the learners have responsibility for their learning activities. However, their control or intentionality is usually constrained in a context where a teacher takes over most of the responsibility for designing learning materials, curricula, the structure of group work, and goals to accomplish. (p. 4)

contrasting CoL and KB
interesting contrasts in the Japanese context (p. 4)

Japanese elementary science lessons fit this type of classroom environment. In the knowledge-building community, on the contrary, participants need to have more responsibility for their own activities and the design of their learning conditions in order to advance their understanding by themselves. They need to regularly engage in objectifying knowledge to be improvable and shared, and they need to use that knowledge to create new knowledge. Participants in a knowledge-building community are, therefore, required to learn strategies not only to understand given knowledge, but also to advance knowledge by themselves. (p. 4)

Toward the knowledge-building classroom (p. 4)

interesting re-engineering of the KB principles — leading to two design principles for Japanese classrooms (p. 4)

By referring to these 12 determinants of knowledge-building, we created two practical design principles. (p. 4)

The first principle was that continuously improvable student ideas are centered in the learning practice. Determinants such as Breal ideas,’’ Bauthentic problems,’’ and Bimprovable ideas’’ were the most crucial issues that we found when designing lessons; our first principle is related to this realization. (p. 4)

Students are told by teachers to raise their ideas at some point, but this activity structure is not primarily designed for students to revisit their ideas for knowledge-building purposes. (p. 4)

Our second principle was that students should manage their ideas from diverse points of view and collaboratively advance their collective knowledge. This principle is related to determinants such as Bidea diversity,’’ Bcommunity knowledge,’’ Bcollective responsibility,’’ and Bsymmetric knowledge advancement.’’ In ordinary Japanese classrooms, the idea of diversity is a quite familiar issue. Students raise many ideas and opinions from their individual points of view. However, their diverse ideas are not transformed into super-ordinate ideas through collective and symmetric activities. (p. 4)

To improve student collaboration in the classroom, we applied our second principle to designing the participatory structure of student activities. Japanese activity structures are normally comprised of whole-class discussion and small-group work. We considered an intermediate level of the participatory structure: inter-group work. Inter-group work is an activity structure where students from different small groups share their ideas and comment on them in a way that bridges the whole classroom talk and the small group work. (p. 5)

Knowledge Forum\ as a knowledge medium for facilitating the knowledge-building practice (p. 5)

Web Knowledge Forum\ (p. 5)

There are three reasons that Web Knowledge Forum\ is a powerful medium. First, learners report their ideas and thoughts in notes; each note is represented as a formatted report, as shown in Figure 1 in the next section. When creating a new note or editing a previous note, learners can also add pictures or movies in HTML format from their private or public directories. Furthermore, they can add links by inputting note numbers. (p. 5)

Web Knowledge Forum\ adds two types of linking information on the note that are mirror images of each other: (1) references, and (2) notes that refer to the original note. The references are a hyperlinked list of notes referred to by the original note. The notes that refer to the original note are a hyperlinked list of notes that refer to the original note. (p. 5)

role of tech
an example of distributed cognition — the tech is enabling a higher level cognition by linking notes together. (p. 5)

Second, notes are reported in the space called Bview.’’ (p. 5)

The structure of views are dynamically created and refined as learning progresses. (p. 5)

role of teacher
an example of the teacher doing higher-level structuring of the knowledge space. Are there concrete examples of students doing this? (p. 5)

The third reason that Web Knowledge Forum\ is a powerful medium is that the administrator can easily order or arrange views, linking one with another or restructuring them. She can also create a view map on the learner’s initial log-in page. (p. 5)

Design studies in Japanese elementary science (p. 6)

Participating classrooms (p. 6)

Science teachers in the school have been involved in our design studies, and we have developed several lesson plans (two lessons a year) through discussion before, during, and after the classroom practices (Oshima et al., 2003). (p. 6)

The classrooms reported on in this study were a sixth-grade class for design study 1 on BAir and how things burn,’’ and a fifth-grade class for design study 2 on BHow matter dissolves.’’ There were 41 students in the sixth-grade classroom and 34 students in the fifth-grade classroom. The lesson on BAir and how things burn’’ continued for 42 class hours (one class hour is 45 min long) in about two months, and the lesson on BHow matter dissolves’’ lasted for 30 class hours in four months due to the inclusion of the winter break. The same teacher, who has more than ten years of teaching experience, was in charge of both classes. (p. 7)

Design study 1: BAir and how things burn’’ (p. 7)

Elicit student ideas or opinions (p. 7)

After the teacher demonstrated that the dense block of newspaper does not burn (or burn very well), the students were asked what is needed for things to burn. (p. 7)

The learning goal for the students in this lesson was collaborative theory construction through experimentation on the burning phenomenon. (p. 7)

Plan investigations (p. 7)

Based on similarities of individual student explanations reported in the form of models (drawings) on Knowledge Forum\, students were grouped into small research teams, each of which pursued their own inquiry into the target phenomenon.1 (p. 7)

Conduct investigation (p. 8)

Each team conducted their experiment by themselves under the supervision of the main teacher. Before their experiments, students were instructed to consider what to observe and record for sharing information with other teams. (p. 8)

Students shared their experiment reports with other teams on Knowledge Forum\ and discussed with the whole classroom how they further advanced their learning. (p. 8)

students had the opportunity to compare varying explanations of the phenomenon under study, and to consider more articulate and convincing theories. (p. 8)

Our contributions to the design of the lesson, based on our two design principles, were: (1) to use students’ explanatory models and experimental reports as conceptual artifacts centered in their science activities, and (2) to get students to engage in collaborative work on their artifacts in order to advance their collective knowledge. Thus, we designed the lesson as sequences of scientific inquiry by small research teams that frequently shared their thoughts and findings on and off line. (p. 8)

Design study 1: Evaluations (p. 8)

Although the access logs provided us with limited information about on-line student activities, they did give us an opportunity to gain insight on how student activities in the lesson were idea-centered. (p. 9)

Notes created by students were first categorized into ideaand factbased. When students drew models or discussed their own or others’ models in notes, the notes were categorized as idea-based. Other notes in which students reported the results of their experiments or experimental procedures were categorized as fact-based. (p. 9)

A t-test on the proportions of ideaand fact-based notes showed that students read significantly more fact-based notes (the mean was 20.10% with 9.05 as SD) than idea-based notes (the mean was 5.70% with 1.78 as SD; t(10) = 5.54, p < 0.01). (p. 9)

However, students focused their attention on constructing their theories within their teams but did not consider their contribution to collective knowledge in the classroom. (p. 9)

Unfortunately, this idea did not get the attention of the other teams. (p. 9)

We concluded from our analysis and observation that our design effort did not satisfy the Bcommunity knowledge,’’ ’’collective responsibility,’’ and Bsymmetric knowledge advancement’’ determinants of knowledge building even though the class could invent models and experimental reports and use them as shared conceptual artifacts. (p. 9)

important recognition
important to recognize some principles were not satisfied, showing challenges for implementing them. it also shows that people could start from somewhere and read other principles as long as the process is iterative. (p. 9)

Design study 2: BHow matter dissolves’’ (p. 10)

In design study 1, we set a target phenomenon for students to continuously engage with through the improvement of their explanatory models. Students engaged in their real ideas in the lesson, but the task itself was not authentic enough for them to compare or synthesize their ideas between small research teams. (p. 10)

The participatory structure in design study 1 was not organized to support students’ engagement in collective knowledge advancement. Collective activity for students to socialize their knowledge in a more global community, e.g., from ideas within a research team to those among teams, and from ideas among teams to those in the classroom as a whole, was implemented in a quite limited part of the total learning process. (p. 10)

Activities were mainly conducted under the teacher’s supervision in classroom talk after students were given opportunities to read and comment on the reports of others in Knowledge Forum. (p. 10)

As the log analysis showed, students were concerned with facts or findings by other teams rather than the ideas of others teams. (p. 10)

In design study 2, the lesson started with the teacher’s question on how students define dissolution. (p. 10)

Elicit student ideas and opinions (p. 11)

The task structure applied in design study 2 was crucially different from that in design study 1. (p. 11)

in design study 2, we asked students to use their conceptual understanding to solve an authentic task— creating a big and beautiful aluminum crystal—and improve their conceptual models through investigation. Since they shared an articulated task goal, the different research teams were expected to engage in more collective and symmetric knowledge advancement. (p. 11)

Exchange information from investigations (p. 11)

In design study 2, we revised the activity structure as follows. First, before students shared information among different research teams on Knowledge Forum\, the teacher encouraged students to briefly report their progress in the classroom talk. (p. 11)

Then, students went back to their research teams to read and comment on the reports of others, and to discuss how they could build new ideas from the reports. (p. 12)

Second, as a result of the classroom talk, the students determined that they needed a control condition in each research team to rigorously test their predictions. They collaboratively designed experiments by distributing different factors for the teams to investigate. (p. 12)

Design study 2: Evaluation (p. 12)

A 2 (Design Study) 2 (Note Type) ANOVA on proportions of accessed notes showed that: (1) proportions of accessed notes in design study 2 were significantly higher than those in design study 1 (F(1, 18) = 16.11, p < 0.01), and (2) proportions of accessed fact-based notes were significantly higher than those of accessed idea-based notes (F(1, 18) = 19.89, p < 0.01) (Fig. 2). (p. 12)

First, student activities in design study 2 were more based on collective and symmetric knowledge advancement. (p. 12)

we infer that student activities in design study 2 were more idea-centered. (p. 13)

The description of students’ activities in team A suggests that the students constantly engaged in improving their ideas through their collaboration with other teams. Activities to systematically analyze and organize information from different investigations that were seen in team A were found in other teams as well. (p. 15)

Discussion (p. 15)

There have been several studies on knowledge-building approaches to science education in other countries (e.g., Hakkarainen & Sintonen, 2002; Lee, Chan, & van Aalst, 2006). Hakkarainen and Sintonen (2002) proposed a new approach to define student scientific inquiry based on the interrogative model (Hintikka, 1988). They succeeded in articulating the process of student inquiry on CSILE (the former version of Knowledge Froum). Lee et al. (2006) investigated the effectiveness of knowledge-building scaffolding for assessing the progress of high school students’ scientific inquiry. They found that the portfolio guided by knowledge-building principles (Scardamalia, 2002) was a powerful tool for high-school students to elevate their level of conceptual understanding of complex scientific concepts. (p. 15)

The focus of our research here was on whether we can refine the current culturally established practice of scientific inquiry by elementary-school students by inventing general but powerful design elements (the task structure and the participatory structure) with a CSCL technology. (p. 15)

In applying Japanese elementary science activity structures in our design studies, we developed two design principles for transforming the class structure into knowledge-building practice: (1) idea-centered lesson, and (2) collective knowledge advancement. (p. 16)

Our observation analysis of student activities on and off line showed that students were involved in scientific inquiry with their ideas being centered in both lessons. They expressed, revisited, and revised their explanatory models through their investigations. (p. 16)

Unfortunately, this emergent problem was not further pursued in their learning. We considered several reasons for students to have missed the important opportunity to deepen their conceptual understanding. One reason is that it was difficult for students to plan and conduct investigations on this issue. The teacher agreed with us that students would not have the repertoire of experimental designs in their minds even if they had been concerned with this problem. The most crucial reason, we believe, is that the worth of the concern was not collectively recognized by other students. (p. 16)

Such an asymmetric or non-collective activity structure kept students from rising above their ideas to form a new perspective. (p. 16)

In design study 2, knowledge advancement was more collective and symmetric than that in design study 1. One reason for such collective and symmetric knowledge advancement is the refinement of the task structure in the lesson. Students were engaged in fun and authentic problem solving (p. 16)

Another design element we can count on is the participatory structure we designed for sharing information from investigations by others. (p. 16)

We found, however, that the implementation of such a new knowledge medium and preparation time for using it did not encourage students to (p. 16)

role of tech
socio-cultural setup greatly influence use of tech. (p. 16)

engage in inter-group communication. (p. 17)

The blending of offand on-line communication for student progress helped them understand what their class as a community knew and what problems or questions remained, or which groups had similar interests and important data. (p. 17)

It may be useful for us to return to the lesson in design study 1 for considering how we can improve the overall lesson practice based on our findings in design study 2. With regard to the participatory structure, we think that we can similarly apply the blending of offand on-line communication depending on the task structure and student activity structure. We need to pay more attention to the task structure, however. Combustion itself is still a mysterious concept that requires further scientific endeavor. It is difficult for us to provide students with a task structure, based on which students themselves engage in authentic and collaborative problem solving, by explaining the scientific mechanism. (p. 17)

We need further collaboration with scientists and curriculum designers to develop a task structure effective for studying combustion in the elementary school. (p. 17)

An issue we have to further consider is that students easily focus on task goals when they require them to do something concrete such as construct a product. (p. 17)

The coordination of doing scientific inquiry and building knowledge through such practices should be further discussed in order to find general knowledge-building activity structures. (p. 17)

Linn, M. C., Lewis, C., Tsuchida, I., & Songer, N. B. (2000). Beyond fourth-grade science: Why do US and Japanese students diverge? Educational Researcher, 29(3), 4–14. (p. 18)

Blog Logo

Bodong Chen



Crisscross Landscapes

Bodong Chen, University of Minnesota

Back to Home