Bodong Chen

Crisscross Landscapes

Integrating Knowledge Building with Open Educational Resources


Open Educational Resources

Inspired by the MIT OpenCourseWare (OCW) which was launched in 2002, the open educational resources (OER) movement, which strives to “champion the sharing of knowledge worldwide to increase human intellectual capacity,” has rapidly expanded throughout the world and is becoming increasingly powerful in the era of knowledge creation and lifelong learning. Most OER projects to date have been concentrated on making quality educational content openly and freely accessible on the Internet. The MIT OCW, which has opened course materials, such as syllabus, course calendar, lecture notes, assignments, exams, labs and projects, hyper-books, simulations, and video lectures, of almost every course taught at MIT, might be a best representative of those projects. A number of other prestigious universities around the globe have also launched their own projects to make various types of course materials open to the public, in their own distinctive ways.

Because of the emphasis on educational content of many projects, a misconception that OER means open educational content or open courseware widely exists. However, improving teaching and learning merely by making course materials openly accessible seems to be a myth. Taking the MIT OCW for example, according to its site statistics of the MIT OCW in 2005 (MIT OpenCourseWare, 2006), 51% of all visits to the site are single page-views, 70% of visitors to the site visit only once, and the average duration of a visit to the site is 11 minutes (without counting all those single-page views). These data that reflect a low use of static course content in the MIT OCW imply a gap of learning support in many OER projects. Furthermore, even if it achieves critical improvements on those analytics, how much a learner learn by reading course materials is still unclear.

As agreed on by several definitions of OER, content is only one aspect of OER. According to the definition Marshall Smith gives in his paper published on Science, open educational resources are content (courses, books, lesson plans, articles, etc.), tools (virtual laboratories, simulations, and games), and software that support learning and educational practice. Candace Thille also notes in her chapter in the book Opening Up Education, although these OER projects have enabled a great step forward in democratizing access to educational materials, the next step of the OER movement is to make instruction, not merely materials, accessible to learners, and improve teaching and learning. According to a report from the William and Flora Hewlett Foundation, “open participatory learning” will be a promising new culture of learning, given such a vast ocean of open educational content created in the past ten years by OER projects across countries. So one important area that is critical for the ultimate sustainability and impact of OER on teaching and learning is about the extent of use of OER. Thus, we should not put our emphasis solely on high-quality open content, but move forward to understand and stimulate use of those open educational materials for the sake of learning. For the next step, it seems urgent to initiate an awful amount of research on how to effectively use open educational materials to really promote teaching and learning.

Open Learning Initiative

Carnegie Mellon Open Learning Initiative (OLI) distinguishes itself from many other content-oriented projects by developing computer-assisted feedback loops for deep learning in its learning environments. With a goal to develop a web-based learning environment that enactments powerful instruction, the OLI compiled a team of content experts, learning scientists, and software engineers. The team’s efforts span from developing better resources, designing cycles of feedback, to conducting studies to deepen our understanding of learning. Among these efforts, developing cycles of feedback is regarded as the pivotal aspect, or the “killer app,” of achieving enactment of instruction in OLI learning environments. The Digital Dashboard for Learning (DDL) is developed for this purpose; it collects learning data as students work through online instructional activities, summarizes this data, and displays meaningful indicators of learning progress to teachers or students. Several evaluation studies reported equivalent learning gains, “accelerated learning,” and better learning retention of standalone learning in OLI than in traditional classroom learning environments. 

However, although OLI has its advantage on formative feedback loops and strong background in intelligent tutoring system comparing with many other projects, a large portion of learning in OLI is expected to happen in the interactions between a learner and well-designed course materials with multimedia, interactive simulation, and feedback tools. In this sense, OLI adopts a didactic pedagogy and does not support peer collaboration in its web-based learning environments. Although individuals now have access to educational opportunities formerly available only to students in top-flight universities, unless they are enrolled in some organized program, they do not have the benefit of social supports, knowledge sharing, and collaborative building of understanding that can be gained in a quality campus milieu. Apparently, the current pedagogical models in OLI as well as many other projects lack affordances that directly address the socio-cognitive dimensions of learning. As argued in the OLCOS Roadmap 2012, since the knowledge society is built to a large degree on digital environments of work and social communication, such (open educational) practices must foster a creative and collaborative engagement of learners with digital content, tools, and services in the learning process.

In the last few years, the use of social software tools and services such as Weblogs, Wikis, social networking, and social bookmarking services beyond the educational sector has played an active and important role in OER. As the OLCOS Roadmap 2012 notes, the success of social software with many users has inspired the idea of a Personal Learning Environment (PLE), which integrates several tools and services a learner can benefit from in self-directed and collaborative online learning. In 2006, MIT OCW tried to launch a customized version of Elgg, an open source social network tool, to allow learners to develop their own portfolios and to gain peer group support through the networking functionality. For now MIT OCW is using a social learning network, OpenStudy, to organize study groups around courses. However, because of an overwhelmingly large portion of peripheral participation in most study groups, level of collaboration in OpenStudy, which mainly happens in the forms of asking and answering questions, is pretty low. The Peer 2 Peer University (P2PU) project has created another model of organizing learning around open educational materials. It runs courses in a six-week semester format. Course instructors or facilitators are all volunteers. Tools for communication and collaboration in a course are usually picked from a variety of tools suggested in the P2PU community course toolbox. Apparently, P2PU goes much further than a peer-tutoring system or a social learning network, in facilitating deeper and more sustaining collaboration, promoting deeper learning, and pursuing ways of accreditation for learning achievement in the OER context. However, the current model of P2PU greatly relies on a lot of volunteering organization and facilitation, and learning is framed in a term-based course running manner. So there is a call for an inclusive learning environment that can weave learning happening on extensive content together for an individual, can support communities of learners, can encourage closer and more sustainable peer-to-peer communication and collaboration, and can enhance socio-cognitive gains in learning with OER.

Knowledge Building

Providing the infrastructure to support peer-to-peer collaboration and discourse is where Knowledge Building theory and technology can play a significant role in the world of OER. Knowledge Building, coined in 1980s by Marlene Scardamalia and Carl Bereiter, is defined as

the production and continual improvement of ideas of value to a community, through means that increase the likelihood that what the community accomplishes will be greater than the sum of individual contributions and part of broader cultural efforts.

Knowledge Building is distinct from learning in that it results in the creation or modification of public knowledge, while learning is an internal, unobservable process that results in changes of belief, attitude, or skill of an individual; personal learning is regarded as a by-product of knowledge building in this theory. Foundational to Knowledge Building theory, as applied to educational contexts, is the notion that ideas ought to be at the center of educational endeavours. These ideas are continually improved through a social process, with a shared goal to advance the frontiers of knowledge as members of the community perceive them.

These educational perspectives has informed the design of Knowledge Forum, software that supports Knowledge Building pedagogy and the process of knowledge creation. Current features in Knowledge Forum have produced strong educational results and have allowed us to identify levels of competence that young students are capable of but are obscured by traditional learning environments and not revealed by current assessments. Knowledge Forum emphasizes embedded and transformative assessment as an integral component of discourse. By collecting and analyzing data from students’ knowledge building activities, such as reading, building-on, referencing and creating views, a rich collection of analytic tools discloses to students extensive information about their progress, e.g., individual and group contribution, vocabulary growth, social network, and semantic overlap between each other’s contributions.

Knowledge building theory and technology support a distinctive pedagogy that is based on 12 Knowledge Building principles (introduced in this article), rather than traditionally based on procedures. Different from didactic instruction, project-based learning, problem-based learning, inquiry-based learning, and many other pedagogical perspectives, knowledge building pedagogy puts discourse for idea advancement at the center, rather than tasks. As Scardamalia and Bereiter note,

For decades educators have promoted constructivist ideas among themselves whereas their students have been expected to carry out constructivist activities without access to the constructivist ideas lying behind them.

While traditional procedure-based pedagogy gives teachers a feeling of security, since they can have definite control over classroom activities, knowledge building discourse embraces a kind of uncertainty, which might not be welcomed by some teachers, but may lead to some phenomenal gains. Such gains are revealed in a number of cases, including the case of a kindergarten kite-making project in the Institute of Child Study and a newly reported case about a bees scientific research done by 8-10 years old in Britain.

Bring Knowledge Building into OER

Integrate Knowledge Building with OLI

OER provides a context in which sustained knowledge-building discourse would be highly desirable. The Institute for Knowledge Innovation and Technology (IKIT), where knowledge building research is headquartered, is undertaking with collaborators at Carnegie Mellon University experiments in merging a knowledge-building environment into OLI. The goal of this collaborative project is to bring expertise from two teams together to provide technology that encourages and encourage collaboration at a distance in achieving deeper understanding. This effort aims to go further than the state-of-the-art in OER to enhance both individual learning and group cognition.

This specific integration involves both technological and pedagogical considerations. The two aspects are highly dependent on each other. On the one hand, pedagogical decisions would shed light on technological integrations. For instance, the current OLI, as well as many other OER projects, adopts a traditional knowledge-transmission pedagogy; learners are expected to interact with pre-designed high-quality course material to absorb authoritative knowledge from it. This pedagogical inclination of OLI is now embodied in its current learning environment which puts linearly organized course content at the center. However, the knowledge building pedagogy thinks that knowledge-building discourse should be at the center, discourse would start from authentic problems and focus on real ideas of learners, and authoritative sources are critically used to understand the present state and growing edge of knowledge in the field for the consistent goal of advancing knowledge. Therefore, implementing knowledge-building pedagogy in OLI calls for a radical change on the technological design of the OLI web-based learning environment. On the other hand, technological affordances and constrains would enable or restrain the implementation of pedagogy. The most widely utilized version of Knowledge Forum is a client-based knowledge building environment developed with Java. The main problem of directly using it in the OER context is that Knowledge Forum is a restrictive environment that lacks the highly acclaimed Web 2.0 capacity of communicating with outer spaces by open APIs. This restriction makes the pedagogical intension of making extensive connections between resources, ideas, and people in the OER context almost impossible. Overall, the interplay between pedagogical and technological aspects will be a central issue in integrating Knowledge Building into OER.

For the initial attempt of integrating Knowledge Building into OLI, the engineer team has implemented a new web-based open source version of Knowledge Forum that can be integrated in any web page. The first integration takes a straightforward approach by directly embedding Knowledge Forum as a discourse component at the bottom of each OLI learning module. Although the new webby Knowledge Forum and its integration with OLI could be regarded as an important move from a confined space to a much more open infrastructure, this integration still looks “superficial” and is far from filling the gap of supporting socio-cognitive dimensions of learning in the OER context.

A case study with this new integration will be conducted in the spring of 2011 with a college level statistics course. This pilot will pursue utility and usability evaluation of current design based on qualitative feedbacks from teachers and students, and will look for ways to improve the design in the next iteration of implementation.

Inclusive environments to support Knowledge Building across OER projects

As discussed above, the absence of supports for socio-cognitive dimensions of learning is a main gap for OER. How can we support individual lifelong learners not solely working in MIT OCW, or OLI, but working across different “open educational content repositories”, or even regular web pages? How could we support their learning by building the socio-cultural layer to make connection with peers, and also to make connections between their ideas? How to draw and manage the boundaries of open learner communities? Although there might be many ways to tackle these challenges, this paper tends to focus on knowledge media design as a direction to make progress.

To enable Knowledge Building across a variety of open educational materials, we need to develop an inclusive environment that can operate with any open educational materials. The ideal is that any media object should be treatable as an object of inquiry and knowledge building. To a very limited extent this is done by attaching a discussion board to such diverse objects as blog posts, news articles, and online videos. At worst these become platforms for venting opinions; at best they are platforms for sharing knowledge and giving advice. Rarely if at all are they platforms for sustained knowledge-building discourse. It is not that such discourse is impossible, only that the technology provides no support for it. Most online communication tools adopt a conversation-oriented interaction with participants’ ideas distributed across messages and responses addressing individual and oftentimes the most recent entries. It then becomes difficult for them to see the whole picture of the extended discourse and understand and review group-level progress in the community’s knowledge. As a result, their discourse contributions are often disconnected and redundant, covering the same ground without significant progress.

Although the need of an inclusive knowledge-building environment is obvious, there are several vital challenges we are confronted. One big issue will be about object granularity and format diversity. On what grain level of object should the environment operate? How to achieve interoperability given such diverse content formats and levels of object granularity? The second challenge is about how to nurture communities of learners. How to increase social cohesion in OER learner communities? How to encourage different levels of participation and make them all contributing to knowledge-building discourse? The third challenge is how to make knowledge-building discourse pervasive in learner’s journey in OER. These three main issues have to be tackled in the design of the envisioned inclusive knowledge-building environment for OER.

Several existing tools might shed light on the design of this environment. Cohere is a tool developed in the UK Open University’s Knowledge Media Institute, funded by the William and Flora Hewlett Foundation, as part of the Open Learning Network (OLnet) project to make high quality learning resources and sensemaking tools freely accessible via the Web (Buckingham Shum, 2008). Idea is the fundamental element in Cohere. An idea is usually linked to one or multiple web resources and emerges during the process of web browsing. What Cohere does is to help users create their ideas, to connect and share ideas, and subsequently connect people who share similar interest. Buckingham Shum (2008) also regards Cohere as a Web 2.0 argumentation tool in that it has several affordances that enable flexible linking ideas and visualizing ideas networks. According to its developers, design of this tool was greatly inspired by Knowledge Building theory and the notion that idea should be put at the center and improved continuously. However, the concept of idea in Cohere is different from that in Knowledge Building theory. The focus of Cohere is to extract “ideas” from open educational materials, make argumentation connections between them, and visualize the connections. The linking effort emphasizes argument relationships between ideas, e.g. “responds to”, “is an example of”, and “positive or negative polarity”, rather than further improving these ideas. Its visualization is implemented in the fashion of concept map, which is good at presenting finalized understanding of knowledge rather than supporting continuous improvement of ideas.

Diigo and Google SideWiki are two interesting tools that can achieve interaction with fine-grained web content. Both of them work as browser plugins, similar as Cohere, and can mark up any web page. Diigo enables highlight and annotation on any text information and images, and it can bring discourse anywhere on the web. SideWiki takes a whole web page as the basic object and expects users to contribute and read helpful information alongside any web page, but it also supports comments targeted at selected text on the page. Their design of user’s interaction with web pages reveals rich possibilities in designing inclusive knowledge building environments for learning in OER. They make discourse pervasive in the web sphere, and help users accumulate materials and personal understanding in their personal libraries in the cloud. However, both tools fall short of the important ability to organize ideas in a flexible way and to encourage continuous improvement of ideas. Google SideWiki aggregates all entries written by a user on this user’s Google profile. Diigo stores web page URLs and mark-ups generated by users into personal or group libraries. All entries are organized basically in a linear manner, and users can use tags and lists to organize information for later retrieval. To support sustaining knowledge building discourse, the envisioned system should have the capacity to keep ideas more actively “living” in the space and to enable learners to easily retrieve and collectively improve their ideas. Such capacity may be achieved through some visualization based on knowledge cartography and semantic techniques.

Knowledge Space Visualizer (KSV) developed by Christopher Teplovs is a prototypic system for showing reconstructed representations of discourse-based artifacts and facilitating assessment in light of patterns of interactivity of participants and their ideas. It integrates techniques from Latent Semantic Analysis and Social Network Analysis to infer relationships among ideas generated in online discourse, to visualize idea network in a community, detect knowledge advancements, and provide formative feedback to students and teachers. A tool like KSV would be very useful in the context of OER in hosting and visualizing rich data in the learning process and making sustaining idea advancement possible.

To summarize, issues like object granularity and format diversity pose even bigger challenges to the integration of Knowledge Building in the whole OER context than in the OLI environment. Sustaining knowledge-building discourse across multiple OER projects calls for an inclusive knowledge building environment that can operate with any kind of web object. Figure below depicts a preliminary design, named “OER Backpack”, created by four engineer students under the guidance of IKIT researchers in November 2010. Rather than solving issues discussed above, such design mockups open up a lot of new issues open to be discussed. Because of the limitation of length, this white paper would not further discuss design considerations embedded in this initial design effort.

Final Thoughts

As the triad OER logic model proposed in the 2007 report from Hewitt Foundation highlights, understanding and stimulating usage of OER is one important challenge, besides sponsoring high-quality open content and removing barriers. The lack of learning support and socio-cognitive dimensions of learning are two disturbing gaps in many OER projects. As a theory rooted in socio-cultural constructivism, Knowledge Building theory, which has been applied in extensive contexts ranging from K-12, graduate school, to professional development in healthcare, could contribute in a way to fill these gaps in OER. However, this integration between Knowledge Building and OER calls for highly sophisticated designs of knowledge media to tackle various issues. This essay specifically explores possible designs to integrate the Open Learning Initiative (OLI) project with Knowledge Building/Knowledge Forum. The ultimate goal is to extend from OLI, which is a single OER project, to the broader OER context. By analyzing new challenges proposed by the new context, several key questions that should be answered in the designs are put forward. These questions, together with a review of several existing tools, help to initiate the pursuit of an envisioned inclusive knowledge building environment for the OER context.