This assignment is a reflection on a system dynamics modelling project undertaken by a small group. Instruction followed a problem based learning (PBL) approach.
In problem based learning, students learn by solving authentic real world problems and reflecting on their experiences. (Gudzial et al.,1997,p.1)
Hence, this project contained two concurrent processes; group problem solving and individual reflection. In the first phase of the project we used system dynamics modelling to solve a complex systems problem about an antelope population. As we proceeded with this group process, we identified and recorded knowledge change and process nuggets, categorised by a common taxonomy. The second phase was an individual analysis of these nuggets to produce this assignment for assessment. This analysis involves meta-cognitive processes that explain how learning has taken place. Through this process the learning experience is made explicit and a deeper understanding can be gained.
In preparation for this project we were given a tutorial that took us step-by-step through the process of building our first system dynamics model. This gave us experience with modelling by studying human population growth. This meant that we could bring what we had learnt in the human case to bear on the antelope case. This approach to study is referred to as case-based reasoning (CBR) (Gudzial et al.,1997).
A facilitator lead us through the initial case study and guided group development of concepts and skills by helping us to identify strategies needed to solve the problem, re-directing our misconceptions, making suggestions on how to manage the collaborative process and by asking pertinent questions.
Problem Based Learning
Our virtual team used synchronous and asynchronous communication tools (chat, wikis and emails) to make group decisions and structure our actions. Communication can be divided into task-oriented and non-task-oriented communication (Veerman & Veldhuis-Diermanse 2001, cited in Schellens & Valcke 2004, p.6). I observed that the synchronous chat was usually about non-task-oriented communication. This included planning and regulating tasks, time management, technical issues, group process management issues such as role allocation, communicating expectations and goal setting. The asynchronous discussion on the wikis was primarily task-oriented. This included presentation of new information, question and answer, explication or evaluation.
The web-based software, Trac, facilitated project management and tracking of participation. Trac allows team members to create milestones that record the team goals with targets for the date of completion. Trac also provides a ticket system that we used to assign tasks to each team member. When the task is completed, the individual closes their ticket. This creates a visible way to keep track of the group progress. Trac also provides a wiki engine for asynchronous discussions and documentation of the task. As a group of collaborating users, we incrementally created a set of linked Wiki pages (Leuf & Cunningham, 2001) enabling us to express, share, and find information compiled by everyone in the group.(Wagner, 2004, p.22).
Our synchronous chat was hosted on the university learning management system, LrnLab.
Systems Thinking and Modelling
Systems thinking is a way of thinking about and describing dynamic relationships that influence the behaviour of systems. (Maani & Cavana, 2000)
Systems thinking is based on the behaviour of feedback and complexity, the innate tendencies of a system that lead to growth or stability over time. (Senge, 2000, P 70) Complex systems can be described by a set of common characteristics: they are open systems that are embedded in other systems (Mc Grath & Tschan, 2004. P 55) and usually contain nested sub-systems, relationships are non-linear incorporating feedback, boundaries are difficult to determine, changes are often hysteretic, and they contain dynamic networks.
We started our study of complex systems looking at Causal-Loop Diagrams (CLDs) and feedback concepts such as reinforcing/positive feedback, balancing/negative feedback and system delays. We then progressed onto stock-and-flow diagram simulations with STELLA software. When creating a STELLA model, “stocks and flows and the inter-relationships between them need to be specified”.(Senge,2000,p.260). The interrelationships need to be specified “in an explicit, mathematical way” (Senge,2000,p.89).
System Dynamics provides us with a very powerful tool for studying the behaviour of complex systems over time. Once a model has been constructed, simulations can be run and model values can be recorded on graphs and tables. The graphs are usually plotted with time on the horizontal axis, and are therefore known as behaviour-over-time graphs. (Fisher,2005)
As I faced the group problem for the first time, I resorted to a familiar tool to try to make sense of it. I tried to model the problem on a spreadsheet. My team members explained to me that this would not take us far because it only offered linear processing of information. It can only make one calculation at a time, although very quickly. STELLA is a dynamic tool that can operate various levels of causality simultaneously.(Mc Grath & Arrow, 2000. P 9).
An obstacle that our group came up against was defining parameters, such as how to split up the pronghorn population into different age stocks. We worked on this issue on and off throughout the project. This may be due to the fuzzy temporal boundaries, which are one of the characteristics of most complex systems.
As we built our dynamic model to include factors such as birth rates, natural death rate, the deaths from coyote predators and the winter cold, our model reached a state of equilibrium (the graph was horizontal). The number of births and deaths balanced each other to give a stable population.
Another concept we grappled with was how the coyote population and pronghorn population affected each other because the pronghorn fawns were the prey for the coyote predator. We can say that the pronghorn and coyote populations were entrained to each other.(Mc Grath & Tschan, 2004. P 55).
Maani and Cavana (2000) suggest that systems thinking and modelling can help us to understand things like the increasing complexity in our lives, the growing interdependence of the world, the revolutions in management theories and practice, the increasing global™ consciousness and yet local™ decision making and the increasing recognition of learning as a key organisational capability.
Design Guidelines for Collaborative Teams
Johnson and Johnson (1996, cited in Schellens & Valcke, 2004, p.4) proposed a number of design guidelines that help to ensure true cooperation will occur in group work. I believe that some but not all of these guidelines were addressed in the design of this project. I think the design guidelines that were not met in the establishment of this collaborative team, impacted negatively on our ability to work effectively as a team.
- Positive interdependence implies that the group shares a common goal, strives after mutual benefit, has a long term perspective and has a shared identity. Our group did share a common goal (to complete the antelope problem). By answering each otherâ€™s questions we strove for the mutual benefit of understanding and the completion of the task. We did not have a long term perspective because the project was only four weeks long and we did not really share one identity because we all came from different backgrounds and different courses.
- Individual and group accountability implies that both the group and each individual will be evaluated. Participation in the group process was going to be monitored but the group product would be evaluated, though not marked. Two members of the team were voluntary participants so they were not accountable. This meant the condition of individual accountability was not met for all members of the group.
- Intertwining face-to-face and CSCL interaction was attempted one week but only three of us turned up. To ensure this condition was met I think this face-to-face participation should have been made compulsory.
- Interpersonal skills relied on written communication. Several of us were foreign students with English as a second language. This would have made communication more difficult for these students. However, I see the presence of foreign students as a positive attribute of the group. Slavin (1987) found that “assigning students of different ethnic backgrounds to work together was consistently related to positive racial attitudes and behaviours, and also reduced the academic achievement gap between minority and majority students.”(cited in Zurita et al, 2005, p.150)
- Group processing was hindered by a limited number of STELLA software licenses. The modelling responsibility fell upon the two team members with the software. The two students doing the modelling did the processing while the rest of us grappled with trying to understand what they were doing. I believe that I would have learnt more about dynamic systems modelling had I constructed the models myself. The cost of this software is a significant issue that may limit its potential for use in public education.
Challenges of Group Work
Lipponen (1999) highlights three difficulties of collaboration that social psychological research has revealed:
- The free-rider effect occurs when an individual member or members of a team off-loads cognitive responsibilities to the other members of the group (Kerr & Bruun 1983; Salomon 1993). Initially, I think we did initially off-load cognitive responsibility onto the two modellers. Once we defined who would check the models, we took back some of this responsibility.
- Status sensitivity occurs when high ability members dominate the group work (Dembo & Mc Auliffe, 1987). I think those of us who had better English writing and communication skills did dominate the discussions but this is inevitable. I think we should have established some turn taking in synchronous chat to ensure everyone had a chance and a responsibility to contribute to the discussions.
- The sucker effect can be seen when an active member of the group tries to avoid the free-rider effect by expending less cognitive effort on team work (Kerr,1983). As I read this I realise that I was reacting to this sucker effect. In the group work last semester I ended up taking on an enormous leadership role in the sense making of an unfamiliar software technology. I purposefully decided that I did not want to be one of the modellers because I did not want the same disproportionate work load I had previously experienced.
Everyone who has participated in teamwork is familiar with these difficulties of collaboration. Collaboration is not always favourable and teams do not always function as best they can (Salomon,1997;Salomon & Globerson,1987,cited in Lipponen,1999)
Academic Controversy in Collaborative Teams
Collaborative work requires bringing together information, ideas, solutions, and opinions that are not always compatible with one another. Each student in the group is expected to present their own position, an argument that states their position in relation to other group members to solve the problem or answer a key question.(Schellens & Valcke, 2006, p. 2)
People involved in a shared task will often experience differences in perspective or differences of opinion. It is essential to establish channels of communication where there is openness and freedom to explore with little risk attached.(Salmon, 2000). If this climate is to be provided, people need to be able to express their ideas without the risk of condemnation. The key to resolving differences is in how we express these differences to one another. “Students were allowed to disagree with one another, but only in ways that respected each others work.”(Hewitt 2004 p. 225) It is through this respect for each other’s work that individuals start to really listen to each other and begin to develop conceptual alignment with each other. The group begins to share common mental models. Full agreement is not necessary to develop collective knowledge but respectful ways of relating to one another is essential.
The two modellers sometimes espoused opposing views on what was the correct way to model. They would come to some form of synthesis (influenced by group discussion or in one case a vote) before we could select the correct model and progress onto the next part of the problem. This displays a dialectical change motor influencing the choice of model. Progress on the models can also be seen as an evolutionary process. The model went through many variations, selecting the best solution or retaining elements we thought were correct. Hence we can consider the development of the model to be driven by dual motors: evolutionary and dialectical.
Peter Senge (2000) describes team learning as something that emerges through group interaction. Through dialogue and discussion, a group develops and transforms their collective thinking. Our group had to learn to mobilise our energies and actions to achieve our goals. A considerable amount of time was taken in the beginning to work out how we were going to do this according to the capabilities of the individual members and the task at hand. Eventually we were able to gather most of our resources to achieve a collective “intelligence and ability greater than the sum of individual members talents” (Senge,2000,p.?).
During the dialogue process, people learn how to think together- not just in the sense of analysing a shared problem or creating new pieces of shared knowledge but in the sense of occupying a collective sensibility, in which the thoughts, emotions, and resulting actions belong not to one individual, but to all of them together.” (Senge,2000,p.75)
The conceptual artefacts we used were in the form of STELLA models, asynchronous discussion, narratives, and graphs. All of these were linked to our central group wiki home page. Many ideas were put forward. If an idea continues to be talked about it is effectively selected by the group and becomes part of the group cognition. If an idea is ignored, forgotten, discarded or publicly relinquished it ceases to be part of the group cognition.
Knowledge Creator Perspective
Wagner (2004) identifies key team attributes that assist the emergence of knowledge:
- The technology supported the changing nature of knowledge (in particular STELLA and wikis). When we became aware of misconceptions or errors in our model we were able to reconstruct the model. If someone could bring insight or new ideas to a discussion they could easily insert their comments at any point in the wikis. Wagner refers to this as dynamically changing knowledge.
- Our asynchronous discussions, prior to making a model, pooled our individual strengths and resources to give the modellers direction. Wagner refers to this as distributed knowledgefor it demonstrates how collective knowledge can be superior to the knowledge of any individual.
- Once the modeller had finished part of the problem, one or two people became the checkers of this model. The checkers sometimes found errors. Wagner discusses this in terms of errors and recovery or quality assurance. Other people can fix the mistakes of individuals.
- Finally, Wagner talks of publication overhead. He suggests that the knowledge content should be the primary concern, not the technology. This relates to the intuitive nature of the user interface. Wikis certainly have low publication overhead because they are so easy to use. STELLA on the other hand has high publication overhead because firstly the cost made it impossible to have a copy for each member and secondly we found it difficult to know when to use the different tools such as reservoirs or conveyors.
When working in collaborative teams the pressure to complete the task can work against the group’s ability to maintain collaboration. Pushing forward may come easily for some of the team but if others are left behind the collaborative spirit is lost. We lost this collaborative spirit at various stages in the project. When this happens it is important for those left behind to ask questions and the leaders to take the time to answer in full. One of the responsibilities of the teacher/facilitator is to pay careful attention to the participation of every member to make sure everyone is feeling part of the team and understands what is going on. Once someone is left behind his or her task-related motivation drops and it becomes difficult to make a worthwhile contribution.
The Learning Culture
Collins & Bielaczyc (1999) suggest that working on a computer allows students to “try things out on their own, without the social ramifications of failure”.(p.134) However, computer supported collaborative learning (CSCL) with PBL, where all communication and work is recorded on a wiki or in chat archives, makes individual contributions highly visible to both peers and tutors. If a student makes a mistake, their peers can see it immediately so the social ramifications of failure can be significant. For this reason it becomes essential to provide a safe environment where everyone can trust that they will not be condemned for their weaknesses, mistakes or failures.
Regardless of how varied the communication between persons may be, it always involves the risk of one person daring to lay him or her self open to the other in the hope of a response. This is the essence of communication and it is the fundamental phenomenon of ethical life. (LĂ¸gstrup,1997,p.17)
Asking for help is an essential part of learning. Some students do not want to reveal that they don’t understand because they are afraid that it will show them in poor light or others will think they are stupid. This sort of behaviour is called learned helplessness (Bruer,1993,p.19,cited in Rick et al.,2006). Attitudes to making mistakes need to be changed so that students focus on developing skills to fix mistakes instead of seeing them as failure. If students believe that intelligence can be improved by effort and will the learning culture will shift towards greater persistence in the face of challenge. (Bransford et al.,2000)
Technical difficulties can obstruct progress in making sense of the STELLA software. The teacher has an important role to assist students with technical issues. Unresolved technical issues can cause students to loose motivation and confidence. This in turn will cause a drop in their level of participation. These students can become passive observers, listening but not activily contributing to discussions. Checking models and writing narratives helps to gain understanding of system dynamics models. An increase in understanding can then help a studentâ€™s confidence so that they are more willing to participate.
Studying complex systems offers a way to generate debate and insight into the dynamics of real world systems. It can help to develop an understanding of the more subtle aspects of nature and behaviour. It has the educational potential to provide ways of combining reductionist approaches to understanding the parts in a system, with constructivist perspectives to understand complex relationships that enable the parts to become wholes.(Maani & Cavana, 2000) It is generally left up to the student to gain understanding in a systemic / holistic way. System dynamics modelling can help us to develop complex systems thinking. Explicit teaching and scaffolding are powerful tools for education. Articulating these complex systems concepts and showing naturally occurring examples to students is a way to encourage deeper understanding and enhance learning experiences.
- Bransford, J., Brown, A. and Cocking, R. (2000) How People Learn: brain, mind, experience, and school. Committee on Developments in the Science of Learning and Committee on Learning Research and Educational Practice, Commission on Behavioral and Social Sciences and Education, National Research Council.Washington, D.C. : National Academy Press, c2000.
- Collins & Bielaczyc (1999)The Enculturalation of Education. Journal of the Learning Sciences Vol. 8. No 1.
- Fisher, D. (2005) Modelling Dynamic Systems: Lessons for a First Course A Study of Population Growth Student Lessons Page 2-3
- Guzdial,M., Hmelo, C., HĂ¼bscher,R., Nagel,K., Newstetter,W., Puntambekar,S., Shabo,A., Turns,J. and Kolodner,J.L. (1997) Integrating and Guiding Collaboration: Lessons Learned in Computer-Supported Collaborative Learning Research at Georgia Tech. Presented at Conference on Computer Supported Collaborative Learning, 1997. retrieved fromhttp://www.oise.utoronto.ca/cscl/papers/guzdial2.pdf 6 Nov 06
- Hewitt, J. (2004)An Exploration of community on a Knowledge Forum Classroom: An Activity System Analysis. In Designing for Virtual Communities in the Service of Learning edited by Barb et al. 2004.
- Leuf, B. and W. Cunningham (2001). The Wiki Way: Quick Collaboration on the Web
- Lipponen, L. (1999) The Challenges for Computer Supported Collaborative Learning in Elementary and Secondary Level: Finnish Perspectives Computer Support for Collaborative Learning 1999 retrieved from http://www.helsinki.fi/science/networkedlearning/texts/lipponen1999.pdf 6 Nov 06
- Maani, K. E., & Cavana, R. Y. (2000). Systems thinking and modelling. Understanding change and complexity. Auckland, NZ: Pearson Education Limited (Textbook)
- Mc Grath, J.E. and Arrow, H., (2000). The Study of Groups: Past, Present, and Future, Personality and Social Psychology Review 2000, Vol. 4, No. 1, 95â€“105
- Mc Grath, J.E. and Tschan, F.(2004) Groups as Complex Action Systems. Chapter from Poole, M. S., & van de Ven, A. (Eds.). (2004). Handbook of organizational change and innovation. Oxford: Oxford University Press.
- Schellens, T & Valcke, V. (2006) Fostering knowledge construction in university students through asynchronous discussion groups Computers & Education, Volume 46, Issue 4, May 2006, Pages 349-370
- Senge, P. M. (2000). Schools that learn: A fifth discipline field book for educators, parents and everyone who cares about education. London: Nicholas Brealey.
- Wagner, C. (2004) Wiki: A technology for conversational Knowledge Management and Group Collaboration. Communications of the Association for Information Systems (Volume13, 2004) 265-289
- Wikipedia resources on Complex systems, Self organisation, reductionism, emergence, agent based model, retrieved 1st Sep 06
- Zurita, G., Nussbaum, M., & Salinas, R. (2005). Dynamic Grouping in Collaborative Learning Supported by Wireless Handhelds. Educational Technology & Society, 8 superior to the knowledge of any individual.
Other Useful Reading
- Poole, M. S., Van de Ven, A., Dooley, K. & Holmes, M. (2000) Organizational Change and Innovation Processes: Theory and Methods for Research. Oxford University Press.
- Jacobson, M.J. and Wilensky, U. (2006) Complex Systems in Education: Scientific and Educational Importance and Implications for the Learning Sciences”