introduction
Collaboration and cooperation have become firmly established as teaching methods in face-to-face classes (e.g., Johnson, Johnson, & Smith, 1998). They are also rapidly becoming widespread in online teaching and learning in both hybrid (mixed traditional and online) course and distance courses. The methods are likely to be most effective if they are firmly grounded in how people actually work together. Some groups collaborate more successfully than others. Frequently, instructors may place students into groups in the expectation that they will collaborate without a clear idea of what collaboration is or how to recognize and encourage it. We must define what we mean by the terms, both so that we can use the techniques successfully and so that we can research them accurately.
In addition, we must distinguish between groups in which people act independently from those who act collaboratively. As Surowiecki (2004) has pointed out, when all the results are aggregated, a large number of people acting independently may give a more accurate solution to a problem than an expert. Interdependent groups may often produce results inferior to the results obtained by their best-performing members or may be affected by a “groupthink” mentality.
Some writers (e.g., Dillenbourg, Baker, Blaye, & O’Malley, 1996) distinguish between cooperation and collaboration. Cooperation, sometimes called “divide-and-conquer,” is defined as individuals in a group dividing the work so that each solves a portion of the problem. Collaboration is the interdependence of the individuals of a group as they share ideas and reach a conclusion or produce a product. If a group of students were given a story to write, its members could cooperate by each being assigned to write a portion of the story and then stitching the parts together. To collaborate, the students would discuss each part of the story, contributing ideas and discussing them until they reach consensus, and then write the story together. Individuals in cooperative groups may compete to produce the best portion of the project. Individuals in collaborative groups cannot compete against one another because they are accountable for the product as a group. Collaborative groups, by definition, share ideas and develop them into new products.
Some instructors use a hybrid technique that involves dividing the class into groups and assigning tasks to be done, with either the students or the instructor choosing the roles. The whole group is then graded on the outcome. Thus, the entire group is accountable for an individual’s efforts, and there is no provision for compensating for a slacker. If one participant fails to complete his or her task, then no one else can step in to complete it. This type of cooperation/collaboration may only provoke resentment and anger and, therefore, it should be avoided.
Collaboration places more challenging demands on individuals than cooperation. Readers actively construct mental representations of text (situational models) to understand situations and make predictions, using a combination of the information in the text and prior knowledge and beliefs (Kintsch, 1994). In a cooperative group, individuals only need to create an adequate situation model of the problem described to submit a solution. In a collaborative group, members must create a situation model and share it with the group. Each must also develop an understanding of the models of other participants so that the group can develop a shared solution.
To study collaboration, we must look closely at the patterns of communication within groups. This is easier to do with text-based online groups than with face-to-face ones, both because there is a permanent record of all interactions and because there are fewer variables in a text-based online discussion (which does not include intonations, facial expressions, and body language). Collaboration cannot occur unless there is roughly equal participation among group participants. Group members must actively respond to one another. If not, they may talk past one another, never reacting or changing as the discussion progresses. The product of the group must be a synthesis of ideas from all the group members. Without these three key characteristics, group interaction may be many things, but it is not collaboration.
online collaboration
The rise of computers and networks has led to new means of computer-mediated communications (CMC). In synchronous CMC, all participants are online at the same time, while asynchronous CMC occurs without time constraints. Synchronous discussion uses chat rooms, instant messengers, or audio and video programs to enable participants to exchange messages in real time. Because of the swift exchange of messages, synchronous discussion may be best suited for brainstorming and sharing ideas. In asynchronous discussions, such as occur over e-mail or threaded Web discussion, students participate at any time and from any location. Participants have more time for considered opinions (Kaye, 1992) and to engage in deeper discussion of ideas (Smith, 1994). Participants are better able to contribute to the discussion equally.
We (Hathorn & Ingram, 2002b; Ingram & Hathorn, 2005), developed analyses specifically for asynchronous CMC using a threaded Web discussion board, where messages are arranged under defined topics, enabling students to add to the discussion with a new message to a series or “thread.” These methods are explained in more detail and demonstrated briefly elsewhere in this topic.
Web-based discussions allow instructors to use instructional strategies in which students solve complex real-world problems. When the groups are successful, learning takes place and students acquire new knowledge and the ability to apply it. The use of text-based messages enables reflection and rethinking of prior knowledge as students ask questions and discuss ideas. In productive discussions, students reflect on ideas while they develop their reasoning abilities through discussion, reading, and analysis (Pugh, 1993).
We have identified three critical attributes of a collaborative group: interdependence (Johnson et Al., 1998), synthesis of information (Kaye, 1992), and independence (Laffey, Tupper, Musser, & Wedman, 1998). These three factors can be used to operationalize the definition of collaboration for research.
The interaction in a group provides insight into how individuals learn through sharing information and testing ideas (Henri, 1992). A key element is the interdependence of the individuals in the group as they work towards the common goal (Kaye, 1992). Positive interdependence leads to individuals promoting learning in others rather than obstructing it (as in a competitive group) or ignoring it (as in a collection of individuals). The individual’s goal cannot be achieved unless the group goal is accomplished (Johnson et al., 1998; Kaye, 1992). Each participant is responsible for contributing to both the other members’ knowledge base and the group project. This process involves offering, challenging, and defending information and experience and making concessions and compromises. Interaction requires participation by all members, responding to one another during the discussion.
The second characteristic of collaboration requires that the group generate a product distinct from the individual contributions of its members. Collaboration must include the creation of new insights during the discussion (Henri, 1992; Kaye, 1992). For collaboration to occur, the group should have a shared goal that requires the synthesis of shared information and ideas. When successful, this creates a product different from any that the individuals could have produced alone.
The third requirement of a collaborative group in education is that the group should be independent of the instructor. This is often difficult for students who are accustomed to referring questions and problems to the teacher rather than using their own resources to find solutions (Laffey et al., 1998). They may try to keep the instructor in the role of knowing all the correct answers instead of developing problem-solving skills with peers (Kaye, 1992). Unless they overcome this tendency, they cannot be a truly collaborative group. Participants and the instructor are all responsible for fostering independence. The instructor must be available for questions but not intrude on the discussion unless collaboration fails.
How can we measure the amount of collaboration in a group, using the three elements of interdependence, synthesis, and independence? It is not possible to use one measure to categorize groups definitively into collaborative or non-collaborative ones. There is a continuum from highly collaborative to barely collaborative groups, so we must measure the relative amounts of interdependence, synthesis, and independence. We can then compare groups for the amount of collaboration they exhibit. In turn, this will allow us to try out various means of increasing collaboration, for example, and to measure the results.
Most Web servers automatically record logs of their activities, so many Web-based discussion boards can automatically track usage statistics (Ingram, 19992000). Analyzing these logs is a common way to measure online participation (Mason, 1992). One can track the dates and times that participants used the boards, the order in which messages are posted, and the threads in which the messages are placed. These data are incomplete, however, because participation does not necessarily indicate collaboration.
Analyzing the threads as entered by the students and generated by the logs may be inaccurate. Students may not place messages and responses in the best places. A message that responds to the ideas in a previous message may be placed in a completely different thread, leaving readers to make the connections on their own. In discussions we have analyzed, the number and length of messages and the apparent links between messages bore little resemblance to the actual structure of the conversation and were not necessarily an accurate indication of collaboration. Participants may communicate and sometimes interact, yet barely collaborate. Statements may repeat or reword previously expressed ideas. The software allowed students to create new discussion threads or to place messages under any specific thread, but participants often used this facility poorly, and the messages were placed almost randomly. If participants assume that others see the situation in the same way, they may make less effort made to convey ideas in a systematic way.
In a second article in this topic, Ingram and Hathorn (2005) demonstrate how to apply these concepts to specific discussions. An important point here is that many of the constructs we are interested in exploring in CMC are complex ones that require multifaceted definitions. They must be studied with reference to the actual communications taking place, instead of, for example, participants’ reactions to the discussions measured through surveys.
future trends
Online collaboration is likely to increase in educational settings as the trends toward group learning and online course activities and materials continue to merge. Some writers (e.g., Weigel, 2002) argue that combining more traditional courses with online collaborations represents a significant step forward in college teaching. Certainly it appears that this can be a very productive marriage of instructional strategy with technology.
Questions remain. How do we make online collaboration work, given the knowledge we have now? Hathorn and Ingram (2002a) suggest that the effectiveness of online collaborations depends on such factors as the task, the technology, the group composition and size, the skills of the moderator, and grading requirements, including individual accountability. Technology skills may play a role, because many undergraduates lack the basic skills needed to use e-mail appropriately in an academic environment. Undergraduates are far less likely to work independently than graduate students.
To increase effective collaboration, we suggest using such strategies as giving the group a meaningful goal, instructing the group to collaborate, providing a process for performing the task, holding the group and its individual members accountable for the result, fostering the interdependence that is so important in collaboration, and giving individuals and groups plenty of practice in collaboration.
There are addition questions that remain to be researched, including which of these strategies are truly effective. Although further work may refine the coding scheme presented here, using a common approach to analyzing online discussions for their collaborative features provides an opportunity to produce research that builds toward solid conclusions. Issues that can be addressed include which factors can be manipulated to produce greater collaboration and whether greater online collaboration results in better problem solving or increased learning.
conclusion
Our goal has been to describe and discuss a method for coding and analyzing online discussions. We have concentrated on asynchronous, text-based discussions in which the key questions include whether the discussion was collaborative and to what extent. Using this coding scheme allowed us to examine the effects on collaboration of such variables as whether the groups were instructed to collaborate and whether they interacted outside of the online discussion board (Hathorn & Ingram, 2002a). A variety of other questions come to mind:
• Do collaborative groups lead to more learning, as many educators think?
• What are the optimal group sizes for producing collaboration?
• How can we structure tasks and groups for optimal collaboration?
By using uniform analyses of group interactions such as those shown here, we can begin to answer such questions. Other analysis methods are possible, of course, but this one has proven to be useful in this situation. Wide use could lead to a growing understanding of the processes of online collaboration. We also foresee expanding the use of the analyses to other sorts of online discussions, including synchronous ones.
key terms
Asynchronous Discussion: The exchange of information that occurs over a period of time. This discussion method allows for reflection and considered opinions.
Collaboration: Group effort characterized by members of a group working together to complete all aspects of a project, and all members of the group are jointly accountable for the finished product.
Cooperation: Group effort characterized by individuals in a group dividing the work so that each member of the group completes a portion of the project.
Situation Models: The mental representation of text created by readers. It is based on information in the text as well as prior knowledge and experience.
Social Loafing: The tendency for some members of a group to do less work than others in a team task in which they are not individually accountable for the product.
Synchronous Discussion: The immediate exchange of messages in real time. It is best used for sharing ideas and brainstorming.
Web-Based Discussion: An asynchronous discussion in which messages are placed on the Web under defined topics.
