Educational Technology & Society 5 (4) 2002
ISSN 1436-4522

Using Construction Kits, Modeling Tools and System Dynamics Simulations to Support Collaborative Discovery Learning

Marcelo Milrad
School of Mathematics and Systems Engineering
Växjö University
Tel: +46 73 396 95 74



In this paper I present my efforts in exploring new ways for designing innovative pedagogical scenarios to support learning about complex domains. My efforts involve the design of interactive learning environments (ILE) to integrate systems supporting alternative ways of interaction with simulations with an emphasis upon support for shared interaction to mediate social aspects of learning, knowledge construction, reflection and design. I will illustrate a particular ILE that has been developed using new IT approaches and computational tools to foster scientific experimentation, modeling and simulation in distributed and collaborative settings. Furthermore, I will present some preliminary results in which I describe how undergraduate students in Computer Science have been using and interacting with this particular ILE during a course called "Computers and Learning".

Keywords: Interactive simulations, complex systems, innovative design, Computer Supported Collaborative Learning (CSCL)


Since the late 1980s, cognitive scientists, educators and technologists have suggested that learners might develop a deeper understanding of phenomena in the physical and social worlds if they could build and manipulate models of these phenomena (Bransford, et al., 1999). Simulation learning environments are having a profound impact in the way we learn and teach about complex problems, both in the social and in the natural sciences (Repenning et al., 1999). System Dynamics modeling tools such as Stella and Powersim and programs like StarLogo and Agensheets enable users to experiment with complex systems and develop better intuitions about the mechanisms that govern dynamic interactions. These type of interactive tools allows for both learning with models and learning by modeling. How can these modeling and simulation tools be used to facilitate learning about complex domains?

Learning about complex domains require more knowledge than any single person possesses because the knowledge relevant to a problem is usually distributed among learners (Spector & Anderson, 2000). Considerable research has documented a variety of difficulties students have with learning concepts relevant to understanding complex systems in a variety of disciplines (Kozma, 2000, Dörner, 1996). Bringing different and often controversial points of view together to create a shared understanding among these people can lead to new insights, new ideas, and new artifacts. New interactive media that allow learners to contribute to framing and resolving complex problems can extend the power of the individual human mind.

Consistent with recent advances in research on learning and instruction are attempts to provide increasingly meaningful learning experiences. In complex domains such experiences include the ability to construct models in addition to using models for experimentation. Recently, Milrad, Spector and Davidsen (in press) have suggested an approach called “Model Facilitated Learning” (MFL) in combination with instructional design principles. Key aspects of this design framework include the use of modeling tools, construction kits and system dynamics simulations to provide multiple representations to help students develop an understanding of problems in situations that involve many interrelated components, are subject to change over time and often involve ill-defined aspects. MFL distinguishes learning by modeling from learning with models and suggests when and why each approach is most likely to be appropriate. In addition, MFL emphasis the notion of socially-situated learning experiences threads throughout elaborated learning sequences. Here, the notion of socially-situated extends to the idea of collaborative modeling.

Collaborative learning is also a current trend in the educational computing community. A first suggestion of how to support collaboration with modeling tools in “discovery learning” has recently been made by van Joolingen (van Joolingen, 2000; Löhner and Joolingen, 2002). In the construction of models using systems dynamics tools, learners engage in cognitive and social processes that promote collaborative knowledge building. Rouwette, Vennix and Thijsson (2000) argue that a collaborative approach to model and policy design is effective for learning and understanding.

A central question I focus on this paper is how interactive learning environments can be designed to augment the cognitive and social processes of scientific understanding and learning. I discuss design principles for such interactive learning environments that use modeling tools, construction kits and simulations to provide multiple representations to help students understand deep, underlying complex problems (Milrad, 2001). I admit at the outset that the efforts reflected in this paper are primarily conceptual in nature, as is most of the learning research in and about complex systems. Furthermore, I present my efforts in exploring new ways to design alternative interactive learning environments to support learning about complex scientific phenomena. I illustrate a particular ILE that has been implemented following  these principles and which it has been used by undergraduate students in Computer Science during a course in "Computers and Learning". I finalize this paper by extending the discussion to provide some preliminary evidences about how students can interact with each other and with the ILE to socially construct an understanding of a complex phenomena in environmental science within the particular domains of water quality and bio-diversity.


Current Pedagogical approaches to learning with  interactive technologies

Current and emerging trends in education are increasingly moving towards learner-centred approaches (Quintana et al., 1999). In these, learning becomes an active process of discovery and participation based on self-motivation rather than on more passive acquaintance of facts and rules (Sfard, 1998). The role of the teacher is coming more to be seen as mentor or guide, facilitating and playing an essential role in this process. From this perspective, learning can be considered as a dynamic process in which the learner actively "constructs" new knowledge as he or she is engaged and immersed in a learning activity. Furthermore, learners will also build understandings through the collaborative construction of an artifact or shareable product (Papert, 1993).The theory of constructivism is at the core of the movement to shift the center of instruction away from delivery in order to allow the learner to actively direct and choose a personal learning path.

An increasing amount of research has been documenting how new constructivist models may be used to reconceptualise curricula, teaching practices, and learning activities, and to effect significant and rich types of learning gains (Cognition and Technology Group at Vanderbilt, 1997). Many new constructivist models of learning utilize the affordances of new computational and communications technologies as part of learning environments in which learners engage in challenging problems, collaboration and creation of shared interaction  (Dillenbourg, 1999).

Social constructivism, an extension of the constructivist approach,  argues that in addition to most knowledge being an interpretation of personal experiences it is also social in nature: knowledge is jointly constructed in interaction. Recent social constructivist perspectives (Jonassen & Land, 2000; Bransford et al., 1999) regard learning as enculturation, the process by which learners become collaborative meaning-makers among a group defined by common practices, language, use of tools, values, beliefs, and so on.  Social constructivism asserts that a particularly effective way for knowledge-building communities to form and grow is through collaborative activities that involve, not just the exchange of information, but the design and construction of meaningful artifacts.

There has also been a growing body of research on authentic and situated learning environments utilizing the problem-based approach to learning (Barrows, 1985). Problem-based learning (PBL) emphasizes solving authentic problems in authentic contexts. It is an approach where students are given a problem, replete with all the complexities typically found in real world situations, and work collaboratively to develop a solution. Problem-based learning provides students an opportunity to develop skills in problem definition and problem solving, to reflect on their own learning, and develop a deep understanding of the content domain (Spiro et al., 1988). This approach was developed in the fifties for medical education, and has since been used in various subject areas such as business, law, education, architecture and engineering. Most recently, there is a growing interest among educators to use problem-based learning in the K-12 setting, and a growing need for problem-based educational software to facilitate the development of higher order thinking skills via technology.

An underlying assumption of all these approaches is that most effective and meaningful uses of interactive technologies to support learning will not take place if technologies are used in traditional ways. According to Jonassen et al., (2000) meaningful learning will take place when these technologies allow learner to be engaged in the following activities:

  • Knowledge construction
  • Conversation
  • Articulation
  • Collaboration
  • Authenticity
  • Reflection


Learning and Understanding about Complex Domains

Complex domains can be depicted as a collection of inter-related items (e.g., stocks and flows in system dynamics), characterized by internal feedback mechanisms, nonlinearities, delays, and uncertainties (Sterman, 1994). These systems typically exhibit dynamic behaviour, especially in the sense that how they behave has an effect on the structure of the system, perhaps strengthening or changing the feedback mechanisms. This change in internal structure in turn has consequences for how the system will behave in the future (Davidsen, 1996). Such complexity is difficult to understand, especially for newcomers to a complex domain. Complex systems can be found in abundance at many different levels. Economics, environmental science, epidemiology, project management, and training all typically involve complex, dynamic systems.

Learning in complex and ill-structured domains places significant cognitive demands on learners, as appropriately recognized by the medical community. Ill-structured domains include those which do not remain constant over time (especially those which change in non-linear ways and those with internal structural relationships that change in significant ways), those which involve variables and constraints which are not well-defined, and those which are influenced in not easily predictable ways by a number of internal and external factors. Ill-structured problems (problem solving in ill-structured domains is the focus in this context) may also require the integration of several content domains. For example, solutions to environmental complex problems, such as acid rain, may require the application of concepts and principles from math, science, biology and political science (Ford, 1999).

Complex learning requires students to solve complex and ill-structured problems as well as simple problems.. The outcome of this knowledge construction is a mental model. Mental models are complex mental representations, composed of numerous kinds of mental components, including metaphorical, visual-spatial, and structural knowledge that result in runnable models of the phenomena being studied (Jonassen et al., 2000). Understanding how a complex system behaves involves being able to provide causal and structural explanations for observed system behavior, and, further, being able to anticipate and explain changes in those underlying causes and structures that may occur as the system evolves over time. This kind of understanding is not acquired easily nor is it likely to be acquired from observations of either real or simulated behavior (Dörner, 1996).

Feltovich et al., (1996) mention that one of the difficulties that people have in learning and understanding complex domains involves the misunderstanding of situations in which there are multiple, co-occurring processes or dimensions of interaction. In these kind of situations, learners often constrict their understanding to one or a small number of the operative dimensions rather that the many that are pertinent. People seem to prefer single models in learning and understanding. These restricted perspectives are the overextended in ways particular research has shown to be detrimental to learning (Feltovich et al., 1996; Kozma, 2000). Knowledge must instead be used and represented in many ways so that it attains many different meanings in different contexts.


Modeling and Simulation Tools for Learning

The notion of interactive tools for modeling and simulating (Maier & Größler, 2000) is quickly gaining importance as a mean to explore, comprehend, learn  and communicate complex ideas. Students are building and using simulations incorporating both procedural and conceptual objectives and doing so in both discovery and expository learning environments (Alessi, 2000). A dimension of particular interest in the educational use of computer simulations is whether and when one learns by building simulations or by using existing simulations (Spector, 2000).

The key feature of an educational simulation is that it makes use of a model to represent a process, event or phenomenon, which has some learning significance. The learner is able to interact with this representation and the simulation provides intrinsic feedback that the learner can interpret as the basis for further interaction. The existence of an underlying interactive model provides the opportunity for a learner to formulate and test a particular hypothesis about a complex system or even to restructure the system. The underlying model may be mathematical leading to the generation of numerical results, rule-based with the intention of providing feedback on subjective input, or even context-based in that the learner is placed in a context that simulates a real situation.

Jonassen et al. (2000) argue that drill-and-practice technology has turned out to be largely ineffective, and that simulation technology based on constructivist learning principles provides measurable learning advantages. According to Spector (2000), interactive simulations  can be useful to improve learning and decision making in complex domains because they :

  • provide opportunities to formulate and test hypotheses
  • can be used to make explicit the causes for unanticipated results in a complex system,
  • can be used to promote interactions with other learners struggling to understand  the  same phenomena

Regardless of the type or format of the simulation, the overriding purpose for simulating systems remains: to provide a learning environment that supports the learner to develop mental models about the interrelationships of variables; to test the efficacy of these models in explaining or predicting events in a system; and to discover relationships among variables and/or confronting misconceptions. However, the extent to which it is helpful to attempt to use interactive simulations to model reality in too many aspects is less evident (Dowling, 1997). While a number of features of the real world which are thought to be relevant to the learning process can be replicated to a certain extent by computer programs, others cannot, and indeed it may well be that maintaining a distinction between the real and the virtual is an important aspect of the transfer of learning from computer-based environments to the wider world. Frequently, the design of these simulation-based learning environments focuses exclusively on computers and the virtual environments they provide, excluding the physical environment. With the emergence of new technologies, and the continued refinements of existing technologies, design potential has expanded dramatically. What kind of interactions should be cultivated, for which types of learning tasks? How should differences in learning tasks influence the design of interaction strategies?


Interactive Learning Environments for Complex Learning: Design Issues and Conceptual Framework

The idea that new technologies will transform learning practices has not yet led to the collaborative ideal. The task of designing effective computer support along with appropriate pedagogy and social practices is simply much more complex than imagined (Stahl, 2002). How can computers and interactive technologies be used to help learners' better understanding of complex and difficult subject matter, and how can computers aids groups of collaborating learners when they, collectively, are trying to understand such material? 

The learning perspective which informs my thinking is based on principles derived from cognitive psychology, learning theory, and instructional design. The learning perspective I find most appropriate is based on notions derived from situated and problem-based learning (Lave & Wenger, 1990), especially as informed by cognitive flexibility theory (Spiro et al. 1988). Instructional design methods and principles consistent with this learning perspective can be derived from elaboration theory (Reigeluth & Stein, 1983) and cognitive apprenticeship (Collins et al., 1989).

Cognitive Flexibility Theory (CFT) (Spiro et al., 1988) shares with situated and problem-based learning the view that learning is context dependent, with the associated need to provide multiple representations and varied examples so as to promote generalization and abstraction processes. Feltovich et al., (1996) argue that CFT and related approaches can help learners develop advantageous skills for thinking and learning about complex subject matter that are like those that are often available when a group is trying to learn together. As a consequence, learning to support the acquisition of such understanding should be designed so as to promote multiple representations (mental models), to promote appreciation of the underlying complexity of the system, and to promote the ability to interrelate various components of the system. Moreover, learning should be supported with a variety of problems and cases.

Land and Hannafin (1996) point out that researchers and designers need to identify frameworks for analysing, designing, and implementing interactive learning environments that embody and align particular foundations, assumptions, and practices. There is a need for learning activities that stimulate an interest for understanding complex phenomena, challenge current understandings and facilitate experience sharing between learners. Spector, et al., (1999) claim that instructional scientists and designers have not fully understood the socially situated learning perspective and its implications for human learning in and about complex domains. According to this view we lack a well-articulated design framework with sufficient detail to take us from a socially-situated, problem-based, collaborative learning perspective to the design of a particular learning environment for a particular subject domain.

More specifically, I am proposing a general approach which might best be characterized as socially-situated, problem-oriented learning in authentic and collaborative settings. This design framework is based on a experiential, problem-based and decision-based learning perspective. I suggest that the design of interactive environments for supporting learning about complex domains should be guided by:

  • Authentic activities: presenting authentic tasks that conceptualise rather than abstract information and provide real-world, case-based contexts, rather than pre-determined instructional sequences. Learning activities must be anchored in real uses, or it is likely that the result will be knowledge that remains inert;
  • Construction: learners should be constructing artefacts and sharing them with their community;
  • Collaboration: to support collaborative construction of knowledge through social negotiation, as opposed to competition among learners for recognition;
  • Reflection: fostering reflective practice;
  • Situating the context: enables context and content dependent knowledge construction; and,
  • Multi-modal interaction: providing multiple representations of reality, representing the natural complexity of the real world.


Figure 1. A Schematic representation of the proposed design framework


The different components of this design framework are rooted in situated cognition which emphasises the importance of situating thinking with complex contexts. Learners are expected to generate problems to be solved and then, learn, develop and apply relevant knowledge and skills through progressive problem generation, framing and solving. The different learning activities which are designed upon this framework require learners to identify research questions and variables, set hypotheses, build and construct experiments, test results, analyse observations and then refine hypotheses and casual variables accordingly. In the next section I present an example of a project where a complex phenomena has been  studied. The learning activities that have been developed are based upon the ideas described in this framework


Interactive Modeling, Design and Construction: A Study with Undergraduate Students in Computer Science

In this section, I describe a science-based project for undergraduate students in which they have been using  a technologically-rich, inquiry-based, participatory learning environment. This activity  consisted of giving learners the opportunity to understand the behaviour and underlying structure of a complex problem in an ecological system (Ford, 1999). Specifically, undergraduate students used  modeling and simulations tools to construct virtual and real models of water quality problems. Thus, by being active in these processes they gained new insights and perspectives of various complex problems in environmental science.


The Class Content

The project which I describe in this section has been conducted and implemented by undergraduate students in computer science during a course called "Computers and Learning". This is a course which takes place during a 20 weeks period. In this course students have close tie to research, in three different ways. The first is that a large part of the course literature are materials in the form of research reports and articles.  The second is that the teachers and guest lecturers connect their teaching to their own research within this field. The third is that the students learn through an explorative learning approach. In order to approve this course, students have to conduct a final project applying all the knowledge and ideas they acquired during this course. The main goal with this final project is to encourage students to explore new design perspectives while building interactive learning environments. 

This course is part of a program called People, Computers and Work (MDA) at the Blekinge Institute of Technology (BTH). The MDA-program addresses the impact and implications of computers and information society technologies in working life and the  development of better solutions concerning systems development. This program is characterized by two major concepts, ICT use and context. The main focus is on how people use interactive systems in a concrete working context.


The Problem Domain and the Learning Tools

As already indicated, I am interested in designing learning environments in which learners collaboratively construct models of complex phenomena. In such environments learners create models by constructing an external representation of these phenomena, which can be fed into modeling and simulation tools to process and visualize the behavior of the model. Learners should also be able to understand the dynamics of the decision-making process with regard to a complex system while they are introduced to a real problem they need to understand and solve.

In this particular case, the problem domain is related to acid rain and its impact on the fish population of a lake. One interesting line of research being followed out in association with this problem involves connecting the project to include real historical data from the lake the students are studying. Moreover, students are expected to analyze and predict what it will happen in the future with the fish population of the lake based on the given conditions. Even if in this case students are studying computer science and not environmental science, this topic has been chosen having in mind the degree of complexity involved in this environmental problem .

In order to get new insights in this problem domain the students must learn a lot about specific aspects of bio-diversity, how to collect data, how to design scientific instruments using IT tools,  how to interpret data, and how these data can be used  in connection to interactive simulations. In this project, learners have access to a number of interactive tools supporting different aspects of complex learning. The tools are a modeling tool, a construction and programmable kit and a simulation environment which are all open for the students to design and work with. These rich technological environments provide an experimental arena for learning in and about complex domains. Thus, learning can take place for instance, through the process of building a device with sensors and a software tool for collecting and analyzing data, constructing relationships and testing sample hypothesis.

The technological tools that are suggested to the learners are: Model Builder, the LEGO-DACTA Robotics System, the ROBOLABÔ programming language, and Powersim. Model Builder (Quintana et al., 1999) is a Java application that supports learners in building and testing dynamic models of complex systems. Model Builder allows learners to create qualitative models of everyday scientific phenomena. The tool support relationship modeling and casual reasoning processes in ways that are conceptually and procedurally accessible for novice learners. Their casual accounts of scientific phenomena are expressed qualitatively using common sense reasoning that servers as a good launching point for eventual accommodation of more formal, quantitative reasoning.  The heart of the LEGO Robotics system is the RCX, an autonomous LEGO microcomputer that can be programmed using a PC. This device can be connected to different sensors to take input from its environment, to process data, and to control signals and devices involved in different processes.  ROBOLAB is the software for controlling the RCX and is based on LAB VIEW Ô. Powersim is a modeling and simulation development environment for PCs. 


Results: The Students’ Project

In this particular section I illustrate how learning by modeling and learning with models, as suggested by Milrad, Spector & Davidsen (in press), can be combined for the design of meaningful learning activities to support complex learning. In order to promote meaningful learning, I propose that learners begin with some kind of concrete operation, manipulating tangible objects in order to solve specific problems. As these operations are mastered, learners can then progress to more abstract representations and solve increasingly complex problems. Following the design principles of model facilitated learning suggested by Milrad, Spector & Davidsen (in press), learners are challenged to solve a variety of complex problems (See Table 1) according to the three stages of learner development:

1. Problem-Orientation

2. Inquiry-Exploration

3. Policy- Development


Task/Complex thinking  component

Cognitive/Social skills

Learning tools and strategies


Which are the factors that influence the PH level of a lake?

Identifying main ideas


Mental Models
Concept Mapping
Problem Based Learning

Model Builder

Putting the problem in a context. Build a device that can monitor the ph and the temperature of the lake?

Determining criteria Concretizing
Inventing a product
Group discussion


Situated Learning
Inquiry Based Learning

Lego Robotics System Robolab Software

Giving the problem a time perspective and a new context. What will happen with the fish population of the lake in 5 years from now?

Hypothesis formulation
Identifying causal relationships
Group discussion

Casual Loops
Model building
Decision-Based Learning

PowerSim Simulations
PowerSim + Robolab

Table 1. Computational media to support learning about complex domains


The first step in the project consisted of a brainstorm, where all six students discussed the problem domain. The main question to be discussed was: which are the different factors affecting the problem under investigation? Furthermore, the students were interested on how  the quality of water in the lake is influenced by all these factors and its impact on the fish population.  To externalize their understanding of the problem each student constructed first, an individual concept map. Thereafter, all six students joined efforts, so the different concept maps merge into a "collective concept map".

In the next activity, the students create some models of the problem and visualize the result of these models by using Model Builder. It is important to mention that Model Builder is a just a tool for qualitative modeling. The group was divided into two sub-groups. Each of the groups built their own model, which also was tested and simulated. By modeling and simulating with this tool, students could test the different hypothesis they had. In a debriefing session the two groups presented their models and results  to each other. Finally, by working together the two groups produced a more accurate model.

After a number of weeks the students' knowledge about the domain increased and they became more enthusiastic about the project.  They wanted to test their hypotheses in a more realistic environment and during a longer period of time, rather than a couple of minutes just by using a simulation. Thus, they decided to experiment with a real aquarium in which they could explore the phenomena to be studied. By using the Lego-Dacta material, the Robolab programming language and a variety of sensors (temperature, pH, light, etc) they design and built a number of instruments for data collection. Some fish and plants were also placed in the aquarium. All the computational tools mentioned above were used to monitor the quality of the water in the aquarium. The main focus of the discussion among students at this stage of the project were issues related to the design and programming of the  LEGO RCX. During a period of two weeks, the devices they built with the LEGO-Dacta material and the sensors were connected to the aquarium. Thus, during this particular period of time the RCX collected the value of some environmental factors (e.g. ph and temperature) and ROBOLAB was used to process to visualize these data in the computer screen.

In order to be able to infer and to assess what it will happen with the fish population during a five years period, the final activity was to investigate (identifying causal relationships, prediction, etc)  these topics by using EcoSIM. EcoSIM is a simulation tool (see figure 2) built by the author in order to allow students to run a system dynamic simulation model of the impact of the ph and temperature on the fish population of the lake for a particular period of time (years 2001-2006). Students were  presented with a number of cases to solve. They were asked to alternate between making modifications to their models and assessing the impact of those modifications, and to conceive and run experiments as they explored the behavior of their models.They were challenged to predict the results of these cases before running the simulation.


Figure 2. A screen shot of the EcoSIM simulation


Thus, the students were able to compare their predictions with the results of the simulation. The discussions and reflections on these analyses brought the students to a new level of understanding. These situations were recorded and transcripted.  The evaluation methodology used during the whole project was based on qualitative methods, drawing on distributed cognition (Salomon, 1993) and the general approach found in activity theory (Nardi, 1996). Specifically, learners were interviewed prior to exposure to the learning experience with the  interactive learning environment. Furthermore, students were asked to complete a web-based dairy related to their learning activities. The questions posed on this web-based diary were focused on individual and group aspects of interactions with other learners, to determine attitudes about specific technology-based learning capabilities, and to record perceived learning value of those capabilities. During the learning experience and subsequent to the experience with the interactive learning environment, learners were again interviewed and asked similar questions. Changes were analyzed so as to determine whether and how patterns of interaction and attitudes evolve as a result of participation in the continuing construction of the different components of the ILE. My preliminary results (Milrad, 2001) provide some evidence that this design approach is effective in the sense that the learning environment engages learners in solving through interactive modeling, design, and using of system dynamics simulations. All these activities are designed to support the collaborative knowledge building process.

The approach described here has certain advantages and disadvantages. One of the advantages is that peer-to-peer discussion and collaboration are effectively supported. Indeed, most of the learning appears to occur in the small group discussions and not only in any interaction or series of interactions with the simulation models. One of the main drawbacks of this approach is that it is very time consuming and applicable for a particular type of problems.

The different components of the ILE are all problem-based but address different aspects of problem solving activities and behavior. These aspects are related to problems directly associated with a concrete and specific environment, to problems associated with hypothesis formulation in a concrete setting, and then to problems associated with abstraction, generalization, and deep understanding of underlying structural causes for observed model and actual behavior. In this type of situated context, learning occurs naturally as a consequence of the learner recognizing knowledge's practical utility as well as the need to use it in an attempt to interpret, analyze, and solve real-world problems (Barab & Duffy, 2000). Figure 3 gives a few glimpses of the results obtained at the different learning stages while learning by modeling and with models in the particular domain of water quality.


Figure 3. From a concept map to a system dynamic simulation through modeling and construction


Implications and Conclusions

In this paper, I  have presented and examined a science based project that engaged learners learning and solving environmental problems through the process of collaborative knowledge building through interactive modeling, design and construction. One of the purposes of the learning tools used in this study was that students should be supported to actively use tools and concepts in their problem solving. The tools support the students' activities in a direction where the kind of concepts used can be increased and more varying. This helps students to expand their engagement and the possible activities they can participate in.  It was important for the  students to know that the problem scenario and simulation models were based on authentic research data gathered from research studies active at the time the simulation was produced. As Brown, Collins, and Duguid (1989) suggest, the problem posed must be applied in an authentic context. In their final web report  the group of students reflects about their experience in this project in the following way:

"We think it has been a very interesting and rich experience. We learnt a lot about the problem domain  and the different computer programs and systems. From our own perspective, we could realize how knowledge construction and learning can be supported by ICT. We have also noticed how our way of thinking about the topics we were studying evolved over time. We got a deeper understanding of this domain and we begin to realize the power of System Thinking  and what it does mean in the real world."

An important aspect of the design approach presented in this paper is the idea that students will gain deeper knowledge about the problem domain while they discuss how they generate and explore new ideas through the manipulation and use of physical and computational models. One of the main differences of this approach, compared to other design approaches for learning with simulations as suggested by Alessi (2000) and van Joolingen (2000), is the integration of physical and computational media in the interactive learning environment, in the spirit of ubiquitous computing (Ishii & Ullmer, 1997).

My initial findings indicate that this approach was a successful innovation in which the learning of a complex domain occurred through construction, modeling and simulation. Students could realize that collaboration through active experimentation, in this case modelingand  designing with the LEGO construction kit, is also an effective way to learn about scientific phenomena, in addition to what can be learnt by using existing interactive simulations. New learning environments, as the one I described in this paper, provide a new arena for discovery learning and collaboration. Based on the preliminary results presented in this paper I have gained critical insights into the design of interactive learning environments to support learning about complex domains using and building simulations. These aspects include:

  • The importance of being able to represent multiple perspectives of a problem;
  • The need to support learning as a shared, collaborative activity-particularly in the context of bridging these multiple perspectives;
  • The need to support interaction, collaboration and reflection both "around the simulation" as well as "beyond the simulation".
  • The concrete instantiation of students’ understandings into cognitive artifacts (Stahl, 2002) facilitated the development of grounded understandings, not as separate concepts stored in the learner’s brain but as distributed descriptions that were situated across and through their experiences.
  • Artifacts for use in designing activities that promote social and cognitive development

Milrad, Spector and Davidsen (in press) suggest that learning by modeling and learning with models should  be combined for the design of meaningful learning activities to support complex learning. From a design perspective, interactive learning environments can be designed to provide learners with symbolic elements that allows them to develop a common background within their discourse. These symbols become something specific they talk about. Furthermore, activities in these learning environments can engage learners in processes that involve authentic scientific investigation, such as inferring, making predictions, observations, assessing and explanations that give a solid background that support their communication. 

I am planing to continue the development and evaluation of these and new ILEs within the framework of a new European research project we will start during the spring this year with our colleagues from Germany, Portugal, Spain and Chile. As I continue to conduct research and to collect more empirical data, I will gain a richer understanding of the potential of this approach for improving the design of technology-rich contexts for supporting learning about complex domains. I would like to encourage all our colleagues in this area of research to continue this exploration so that we may ground theoretical conjectures regarding the potential of these contexts in empirical findings.More broadly, I hope that further research will help us to develop a richer theoretical framework for understanding the role of these new kind of learning environments for learning about complex domains.



Alessi, S. (2000).Building versus using simulations.In J. M. Spector & T. M. Anderson (Eds.), Integrated and holistic perspectives on learning, instruction and technology: Understanding complexity. Dordrecht: Kluwer.

Barab, S & Duffy, T. (2000) From Practice Fields to Communities of Practice. In Jonassen, D. & Land, S.  Theoretical Foundations of Learning Environments.Lawrence Erlbaum Associates, Publishers.

Barrows, H. S. (1985). How to design a problem-based curriculum for the preclinical years, New York: Springer-Verlag.

Bransford, J. D., Brown, A. L., & Cocking, R. R. (Eds.). (1999). How People Learn: Brain, Mind, Experience, and School. Washington, D.C.: NationalAcademy Press.

Brown, J. S., Collins, A. & Duguid, S. (1989). Situated cognition and the culture of learning. Educational Researcher, 18 (1), 32-42.

Cognition and Technology Group at Vanderbilt. (1997). The Jasper project: Lessons in curriculum, instruction, assessment, and professional development. Mahwah, NJ: Lawrence Erlbaum Associates.

Collins, A., Brown, J. S., & Newman, S. E. (1989). Cognitive apprenticeship: Teaching the crafts of reading, writing, and mathematics. In L. B. Resnick (Ed.), Knowing, learning, and instruction: Essays in honor of Robert Glaser, Hillsdale, NJ: Lawrence Erlbaum Associates, 453-494.

Davidsen, P. I. (1996). Educational features of the system dynamics approach to modeling and simulation. Journal of Structural Learning, 12(4).

Dillenbourg, P. (1999). Collaborative Learning: Cognitive and Computational Approaches. Pergamon.

Dowling, C. (1997). Simulations: New "Worlds" for Learning? Journal of Educational Multimedia and Hypermedia, 6 (3/4), 321-337.

Dörner, D. (1996). The logic of failure: Why things go wrong and what we can do to make them right (Translated by R. Kimber & R. Kimber), Metropolitan Books, New York.

Feltovich, P., Spiro, R., Coulson, R. & Feltovich, J. (1996). Collaboration with and among minds: mastering complexity, individually and in groups.  In CSCL: Theory and Practice of an Emerging Paradigm. Mahwah, NJ: LawrenceErbaum,  Associates.

Ford, A. (1999). Modeling the Environment: An Introduction to System Dynamics Modeling of Environmental Systems,.Island Press.

Ishii, H. & Ullmer, B. (1997). Tangible Bits: Towards Seamless Interfaces between People, Bits and Atom.  In Proceedings of ACM CHI '97, New York: ACM Press, 234-241.

Jonassen, D & Land, S. (2000). Theoretical Foundations of Learning Environments.Lawrence  Erlbaum Associates, Publishers.

Jonassen, D; Peck, K. & Wilson, B. (2000). Learning with Technology.: A Constructivist Approach. Prentice Hall.

Joolingen, W.R., van (2000). Designing for collaborative discovery learning.In G. Gauthier, C. Frasson & K. VanLehn (Eds.).Intelligent Tutoring systems.Berlin: Springer.

Kozma, R. (2000). The Use of Multiple Representations and the Social Construction of Understanding in Chemistry.In Innovations in Science and Mathematics Education: Advanced Designs for Technologies of Learning, edited by Michael J. Jacobson, M and Robert B. Kozma.Lawrence Erlbaum Associates, Publishers.

Land, S & Hannafin, M. (2000).Student-Centered Learning Environments.In Jonassen, D. & Land, S. (Eds.) Theoretical Foundations of Learning Environments.Lawrence Erlbaum Associates, Publishers.

Lave, J. & Wenger, E. (1991). Situated Learning: Legitimate Peripheral Participation.CambridgeUniversity Press.

Löhner, S. & Joolingen, W.R., van (2002).The effect of representations on communication and product during collaborative modeling.Proceedings of CSCL 2002. Hillsdale (NJ): Lawrence Erlbaum.

Maier, F., & Größler, A. (2000). What are we talking about? A taxonomy of computer simulations to support learning. System Dynamics Review, 16 (2), 135–148.

Milrad, M, Spector, M & Davidsen, P. (in Press).Model Facilitated Learning. Book chapter to appear in "E-Learning: Technology and the Development of Learning and Teaching". Kogan Page Publishers UK.

Milrad, M. (2001). From Concrete Experiences to Abstract Formalisms: Learning with Interactive Simulations that Combine Physical and Computational Media. In Okamoto T., Hartley R., Kinshuk & Klus J. (Eds.) Advanced Learning Technology: Issues, Achievements and Challenges, Los Alamitos, CA: IEEE Computer Society, 141-144.

Nardi B.  (1996). Context and consciousness: Activity Theory in Human-Computer Interaction, Cambridge, MA: MIT Press.

Papert, S. (1993). The Children's Machine: Rethinking School in the Age of the Computer. New York: Basic Books.

Quintana, C., Eng, J., Carra, A., Wu, H-S., Soloway, E. (1999): Symphony: A Case study in extending learner-centered design through process space analysis, Proceedings of CHI‘99, 473-480.

Reigeluth, C. M. & Stein, F. (1983).The elaboration theory of instruction. In Reigeluth, C. M. (Ed.), Instructional design theories and models: An overview of their current status, Erlbaum, Mahwah, New Jersey.

Repenning, A.; Ioannidou, A. & Phillips, J. (1999).Collaborative Use & Design of Interactive Simulation Components. In C. M. Hoadley and J. Roschelle (Eds.), Proceedings of the Computer Support for Collaborative Learning (CSCL) 1999 Conference, 475-487.

Rouwette, E. A. J. A., Vennix, J. A. M., Thijssen, C. M. (2000). Group model building: A decision room approach. Simulation & Gaming 31(3).

Salomon, G. (1993). Distributed Cognitions:  Psychological and Educational Considerations.CambridgeUniversity Press.

Sfard, A. (1998). On two metaphors for learning and the dangers of choosing just one. Educational Research, 27(2), 4-12.

Spector, J. M., Davidsen, P. I. & Wasson, B.  (1999). Designing Collaborative Distance Learning Environments for Complex Domains.ED-MEDIA 99 Conference Proceedings.

Spector, J. M. (2000). System dynamics and interactive learning environments: Lessons learned and implications for the future. Simulation & Gaming 31(4).

Spector, M., & Anderson, T. (2000). Integrated and Holistic Perspectives on Learning, Instruction and Technology: Understanding Complexity. Kluwer Academic Publishers.

Spiro, R. J., Coulson, R. L., Feltovich, P. J., & Anderson, D. (1988). Cognitive flexibility theory: Advanced knowledge acquisition in ill-structured domains. In Patel, V. (Ed.), Proceedings of the 10th Annual Conference of the Cognitive Science Society (pp. 375-383). Mahwah, NJ: Erlbaum.

Stahl, G. (2002). Contributions to a theoretical framework for CSCL. Proceedings of CSCL 2002, Hillsdale (NJ): Lawrence Erlbaum, 62-71.

Sterman, J. (1994). Learning in and about complex systems. Systems Dynamics Review 10 (2-3), 291-330.


Copyright message

Copyright by the International Forum of Educational Technology & Society (IFETS). The authors and the forum jointly retain the copyright of the articles. Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear the full citation on the first page. Copyrights for components of this work owned by others than IFETS must be honoured. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from the authors of the articles you wish to copy or