Issue
Article
Vol.30 No.2, April 1998
Article
Issue
Issue |
Article |
Vol.30 No.2, April 1998 |
Article |
Issue |
For learners, conversational interactions are critical [1,9,13]. "Stuff", e.g. printed materials, videos, web pages, just doesn't do it. Learners need to ask for clarification, they need to ask for elaboration, and they need to see information presented in other ways. In short, they need "a conversation".
Is this surprising? Consider rich academic communities -- arguments in the halls, small group seminars; coffee houses are often a place for intellectual discussion, etc. This is the model of learning that we need to take seriously in developing our technologies. Imagine if students could have technologies that let them ask for clarification of ideas quickly and let them participate actively in the construction of their own understanding with constant feedback. If we can scale conversations then we will have really scaled critical learning opportunities.
What will the technology be like that will support learning conversations? How will it be used? How will it fit into our existing educational contexts?
Over the past five years, as part of an Apple research project, we have been trying to answer these questions. Working in collaboration with educators, we have been experimenting with using technology to enhance learning conversations. These experiments have led to the development of a descriptive model of characteristics inherent to situations where learning conversations take place. This model is presented in the next section.
Learning conversations don't just happen. Contexts where learning conversations occur share certain characteristics. We have articulated six characteristics in the model of learning conversations illustrated in figure 1.
Figure 1: A model of Learning Conversations. All six elements need to be in place for learning conversations to be possible
The characteristics of situations conducive to learning conversations are:
Events bring people together and establish a time and a place for conversation. Good events have their own internal structure that helps guide the conversation. An event starts with something that gets the conversation going, introductory remarks or an activity are techniques commonly used. There is then an intermediate period where the bulk of the conversation takes place. Finally, there is a culminating activity, perhaps someone summarizing the conversation, or making future arrangements. Clearly structuring events helps structure the learning experience for the learner.
Not any conversation enhances learning. An exploratory kind of conversational style is particularly effective [6,14]. In a successful learning conversation, participants: generate explanations and produce justifications, are willing to see things in new ways, challenge current understandings, are open to views that conflict with their own, are happy for people to question them; and do not dismiss and routinely criticize other ideas [4]. In fact, conversations that lead to learning tend to take a certain form [2]. Participants extend, paraphrase, refine, complete or critique a partners reasoning. For example some typical phrases are:
- extend
- `I have a further thought...',
- paraphrase
- `I see, what you mean is ...',
- refine
- `I should have made the distinction between... and ... clearer.'
`Here's a further thought ...',- complete
- `Is what you're trying to say...',
`I agree and in addition ... is also going to have an effect ...'- critique
- `Aren't you missing this important point...',
`Don't they distinguish .....'
In a good conversations participants have a mutual interest in the topic of conversation. In addition, participants need to listen to each other, have mutual respect, trust one another, and let each other have equal opportunities to speak and share their thoughts. Perhaps Azmitia and Montgomery's [1] finding that friendship groups often learn better, is precisely because such groups have mutual respect, trust each other, allow each other more equal air time and listen to each other more.
Without access to expertise, learners would have to rediscover everything from scratch! Such expertise is available from secondary sources -- books, videos, etc. -- but learning happens best through direct experience. Learners can successfully experience activities when working with an expert that would be impossible for them on their own [15].
Talking to experts also exposes novices to the different styles of conversation and to specialized vocabulary that occurs in different domains. For example, the talk of professional biochemists differs from that of psychologists, which in turn differs from the kinds of talk found in the field of marketing. Learning the style of talk for a particular language community is an important part of becoming a member of that community [8].
Not all experts are necessarily effective in enabling learning conversations. Experts need to be able to perceive learners misconceptions and produce explanations that are readily accessible.
Conversations are about something. Learning conversations typically happen when someone has arranged for a group of people to come together for the purpose of solving a problem or doing a task. There needs to be a topic of conversation. It's hard to start a conversation when you don't share a topic of interest, or don't know what topic is of common interest. The best learning conversations occur when there is an established topic and participants have had a shared experience that they can draw on to focus and direct their conversation. In addition, direct experience with phenomena is important for learning because it provides concrete examples that learners can talk about even before they understand and have the vocabulary to describe the underlying concepts and mechanisms. Such experience comes through engagement in well designed activities.
In a conversation words tend to get used in a range of meanings far broader than their `canonical' dictionary definitions might imply. This can make it hard for participants to understand one another and establish a common frame of reference [3]. One way to establish a common meaning for the words being used is by pointing to artifacts that make our references clear and disambiguate them. For example, we may point at a particular artifact to illustrate what we mean, draw diagrams on a whiteboard, etc. Interestingly, research has found that participants in text-based on-line conversations have difficulties in establishing a common-ground [10]. Perhaps this is because they are unable to refer to, and ground their conversation in, conversational props. Learning conversations need to be grounded in real-world experiences that learners can relate to. Conversational props help learners refer to and share these experiences.
When one or more of these characteristics is missing from an educational situation, it becomes hard for learning conversations to emerge. For example, if there is no expertise available in the group, then students may never discover important concepts for themselves. If students don't have a learning attitude of asking questions and listening to each others responses, then they will not be able to build on each others understandings. If students can't ground their conversations by sharing conversational props, then they will find it hard to establish a common understanding. If there is no well established rhythm to the conversation, then students will not know when to come together to engage in conversation, and when they can work at their own pace and engage in personal reflection. Such situations are less than optimal for learning conversations.
Technology can be used to enhance many aspects of learning conversations. It can provide tools for creating and sharing conversational props. It can allow learners access to non-local peers and experts. It can provide shared artifacts such a schedulers that help maintain the pacing and rhythm of the conversation. It can provide resources and examples for hand-on activities. In the following section we describe some of these opportunities in more detail. Following this we describe a particular example of technology being used in a real-world context to support learning conversations.
Technology can support and extend learning conversations. For example, the telephone extends conversations beyond the face-to-face to situations where participants are in physically remote locations. Faxes allow rapid exchange of rough graphics around which conversations can take place. Whiteboards are used to create and display text and graphics during conversations between large and small groups. New media tools can be provided in computers that integrate traditional tools such as the telephone and whiteboard, and provide new features. Some possibilities are:
Creating conversational props is not easy. For most people, some skill development and hours of practice is required to become good at sketching or diagraming. However, tools that facilitate image capture can help a lot, cameras can be used to capture still images and video cameras can capture moving images showing process and changes over time. Media databases can also provide a resource for conversational props. Similarly, data modeling and simulation tools can be used to enable students to create visualizations of abstract relationships.
Communications technologies allow us to converse with people around the world, extending the number of people we have access to. Obviously this introduces new problems such as finding appropriate people to talk to. Networks not only allow students to communicate with distant peers [11], but also with experts from around the world [5]. Talking with experts is highly motivating for a learner, however, finding an expert who is both willing and able to talk to a learner can be difficult. Even special programs that have been set-up (for example, Genentech's Access to Excellence) have to carefully manage the interactions between experts and learners to ensure that the experts are not overwhelmed with eager learners asking them questions?
Using networks, people can access vast libraries of on-line information. They can also share their own information artifacts. Networks also allow ready access to specially designed activities. These activities can be regularly updated, and new activities can be made available. Activities may be interactive simulations, etc., or they may be on-line interfaces to real-world artifacts such as robotically controlled landscapes, electron microscopes and telescopes.
Through the use of structured templates of conversational exchange, technology has been used to encourage a learning attitude [5, 7]. For example, learners using the CSILE [12] system have to label their contribution (theory, hypothesis, criticism, etc.) to indicate what role it played in the conversation. This is a very structured approach. More flexible approaches might simply provide tools for differentiating conversational threads as in conventional on-line discussion groups, or provide support for annotation of others contributions.
Video can be used to readily capture conversational props to illustrate points made in a conversation. However, sharing these video illustrations is not easy. We developed the Multimedia Bulletin Board, prototype software that uses the metaphor of a physical bulletin board, to make it straightforward for learners to share video illustrations as part of a learning conversation.
The Multimedia Bulletin Board displays a set of boards, each containing video messages. An example board is shown in figure 2. Users can create new messages and edit existing messages. It also enables the annotation of existing messages, the creation of new boards, and the browsing of different boards full of messages. As the users created boards and messages using this tool, they are distributed electronically for annotation and reply.
Figure 2: An example of a screen from the Multimedia Bulletin Board
In collaboration with the Exploratorium science museum we conducted a trial of this bulletin board in 1993 with a class of fourth graders who were studying weather. Students worked in groups and used a video camera to observe and film real-world weather phenomena (puddles, clouds, etc.). Students created video messages containing their questions about phenomena they had observed. They illustrated these with video they had created. These messages were sent to the Exploratorium, where staff annotated them, or sent back new messages containing answers to students questions in the form of video of relevant museum exhibits or descriptions of simple experiments students could conduct in the classroom. Each modification to the bulletin board provided a context for more student inquiry and requests for clarification or more resources
Figure 2 shows a board developed by students who were studying weather, including questions such as Why do clouds move? and Do clouds disappear? Messages containing movies of relevant museum exhibits, or descriptions of simple experiments such as a demonstration of an experiment that answered the question `why clouds are dark on the bottom'. It showed that if you stack wax blocks up and then shine a light on them, then the blocks at the top will be light but as you go down they get darker because the light diffuses. The students did these experiments for themselves in the classroom.
The teacher often used a projection device to show the boards on a big screen and then led a class discussion about the messages. Students then helped each other understand the experts' messages and jointly formulated new questions in new messages. The use of this presentation capability of the computer encouraged local collaboration and group inquiry to complement the remote collaboration. The following is an example of a learning conversation that happened in forth grade classroom. The students had just viewed the message described above showing why clouds are dark on the bottom. The demonstration shows that light as it passes through the cloud, the light scatters in all directions.
- Student 1:
- If clouds aren't solids or liquids then I don't understand how the light couldn't get through that, because if it's just a gas, then the light should be able to get through the gas.
- Student 2:
- It's not gas, it's liquid, only it's not very heavy liquid it's like bitty, bitty drops.
- Student 3:
- It's like in the form of a gas.
- Student 2:
- Yes, like in the form of a gas.
- Student 1:
- Then why can't it get though?
- Student 3:
- Because water can be a gas too.
In this example, the student 1 is confused, because the message said that the cloud scattered light because it is made of water drops. However, this student thinks that the cloud is a gas. The other students try to correct the implicit assumption in the first students question which is that if a cloud is a gas it can't contain water drops. This example shows some of the features characteristic of a learning conversation. For example, the students ask questions and they identify assumptions underlying the questions being asked.
The kinds of learning conversation shown above were common. Throughout the trial the students got much better at asking questions and listening to each other. All the characteristics that are important for encouraging learning conversations were present. Students collaborated with peers, both in small group and in class discussion. Students had access to hands-on activities to provide a focus for their inquiry and conversation. There were experts who could scaffold the students inquiry and conversation. The project allowed time for individual reflection and time for discussion; and there were events such as times when messages had to be created in order to be sent to the Exploratorium, and messages arriving from the Exploratorium. These events helped students structure their time. Finally, the entire curriculum was focused on having students adopt a learning attitude. They were encouraged to ask questions of themselves and others, and to listen to responses both from their peers and from the experts.
It was not only the students who engaged in learning conversations. The teacher used the bulletin board system to ask her own questions of the Exploratorium experts. For example, she asked them content questions when she was unsure about how to answer students' questions. When she thought students were having difficulty understanding, she asked for classroom activities that might elaborate on a concept. This contact provided the teacher with vital interaction with adult colleagues, thus breaking the isolation typical of teaching and providing immediate training in science in her own context.
Being able to engage in learning conversations is critical for learners. Technology is not a necessary component of such conversations, however, technology can extend these conversations as shown in the example above. In order to support on-line conversations, technologies must facilitate the inclusion and integration of conversational props. Furthermore, we need to design services that distribute compelling activities around which conversations can take place, bring together learners with common interests and include experts who can facilitate learning and encourage appropriate kinds of conversational exchange.
During the last five years at Apple we have produced, and seen others produce, several additional compelling examples of the use of technology to support learning conversations on a small scale. The question is how to take these ideas and extend them so that they are available to more learners. How can we make conversational interactions with scientists available to hundreds, or thousands of students! We certainly can't have them all communicating with the scientists at the Exploratorium!
[1] Azmitia, M. and Montgomery, R. (1993). Friendship, transactive dialogues, and the development of scientific reasoning. Social Development, 2, 202-221.
[2] Berkowitz, M.W. and Gibbs, J.C. (1983). Measuring the development of features of moral discussion. Merill-Palmer Quarterly, 29, 399-410.
[3] Clark, H. H. (1992). Arenas of Language Use. Chicago and CSLI.
[4] Coleman, E.B. (1992). Facilitating conceptual understanding in science: a collaborative explanation-based approach. Doctoral dissertation, University of Toronto.
[5] Edelson, D.C. and O'Neill, D.K. (1994). The coVis Collaboratory Notebook: Computer Support for Scientific Inquiry. Presented at the Annual Meeting of the American Educational Research Association, New Orleans.
[6] Fisher, E. (1992). Characteristics of children's talk at the computer and its relationship to the computer software. Language and Education, 7, 2, 187-215.
[7] Hoadley, C.M. and Hsi, S. (1993). A multimedia interface or knowledge building and collaborative learning. Proceedings of InterCHI'93, International Conference on Computer-Human Interaction , Amsterdam, The Netherlands. ACM.
[8] Lave, J. (1988) Cognition in practice. Cambridge University Press.
[9] Mercer, N (1994) The Quality of Talk in Children's Joint Activity at the Computer. Journal of Computer Assisted Learning, 10, 1, 24-32.
[10] McCarthy, J.C., Miles, V.C. and Monk, A.F. (1991). An experimental study of common ground in text-based communication. Proceedings of CHI'91, Conference on Computer-Human Interaction, 209-215.
[11] National Geographic Society. (1988). Kids Network. Washington DC: National Geographic Society.
[12] Scardimalia, M. and Bereiter, C. (1993). Technologies for Knowledge-Building Discourse. Communications of the ACM, 36, No.5, 37-41.
[13]Teasley, S. D. (1995) The Role of Talk in Children's Peer Collaborations. Developmental Psychology, 31, 2. 207-220.
[14] Wegerif, R. (1996) Using computers to help coach exploratory talk across the curriculum. Computers and Education. 26, 1-3, 51-60.
[15] Vygotsky, L. S. (1978) Mind in Society: The Development of Higher Mental Processes. Cambridge: Harvard University Press.
Rachel Bellamy has conducted research in user interface design since 1987. For the last five years, she has worked in Apple's Advanced Technology group. Her work at Apple has focused on the design and use of collaborative technologies to support K-12 and Organizational Learning.
Kristina Woolsey is a distinguished scientist at Apple Computer. Previously she managed the Apple Multimedia Lab, and has produced several award winning multimedia titles for education.
Issue |
Article |
Vol.30 No.2, April 1998 |
Article |
Issue |