Computer simulations and scientific knowledge

Draft provided only for reference – Jimoyiannis A. (2008). Computer simulations and scientific knowledge construction. In A. Cartelli & M. Palma (Eds.), Encyclopedia of Information Communication Technology (pp. 106-120). Hershey, PA: IGI Global.

Computer simulations and scientific knowledge construction Athanassios Jimoyiannis Department of Social and Educational Policy University of Peloponnese, Greece [email protected] INTRODUCTION Information and Communication Technologies (ICT) is increasingly expanded nowadays and undoubtedly constitutes a vital component of our modern society influencing many aspects of our lives, such as administration, economy, culture, work environment, home-life and most of all education. In particular, multimedia and Internet technologies provide exciting opportunities for the integration of new tools in the curriculum in order to support teaching, to promote students’ active engagement and enhance their ability to facilitate high order skills. A number of ICT applications, such as computer-based laboratories, hypermedia and virtual reality applications, educational games, simulations and modelling tools, exploratory programming environments, intelligent tutors and others are available for teachers and students (Jonassen et. al., 2003). Among the various ICT applications, computer simulations are of a great interest since they constitute open educational environments providing active engagement and practical experiences for learning and the understanding of concepts beyond their theoretical context. Currently the use of simulations cover a wide range of applications within the areas of research and analysis studies (Feinstein & Park, 2002; Hanan et al., 2002; Mesa et al. 2003; Washington et al., 2000), system design (Axelrod, 1997; Lorek & Sonneschein, 1999), training and education (Ziv et al., 2000; de Jong & Joolingen, 1998; Jimoyiannis & Komis, 2001, Lee et al., 2004), entertainment (Leemkuil et al., 2000) and physical therapy (Merians et al., 2002). Computer simulations are becoming more generally recognised as efficient learning environments where students can explore, experiment, question and hypothesise about real life situations (natural or social), which would be inaccessible otherwise. Simulations can offer substantial benefits in education by overcoming obstacles on doing experiments, through replacing real world systems, overcoming drawbacks of those systems, visualizing invisible processes and offering multiple views and multiple representations of the situated system. In this article the basic characteristics of scientific and educational simulations are discussed. Research findings which support their educational effectiveness are presented and, emphasis is placed on the pedagogical issues of designing and using simulation environments aiming at facilitating students’ engagement and active knowledge construction. SCIENTIFIC SIMULATIONS Generally speaking, a simulation is a technique of imitating the behaviour of a situation, process or system by means of an analogous system. In the simplest sense a system is a set of interacting identities. In the case of scientific simulations this analogous system is a mathematical model. The mathematical equations that produce the model represent the various processes which take place within the target system. In other words this model constitutes a simplified or idealised representation of a system by means of a set of mathematical equations (algebraic, differential or

1


integral). The mathematical model becomes a simulation by solving numerically (i.e. for varying sets of input values) the equations comprising in order to imitate or simulate the dynamic (time-varying) behaviour of the system (Fishwick, 1995). In a computer simulation the mathematical model is produced by proper executable algorithms, which are used to solve the mathematical equations. Consequently, a computer-based simulation is a software application that embodies a model of the actual or theoretical system, executing the model on a computer and analysing the output. Any system in either the micro- or the macro-world can be simulated, providing that its behaviour can be described by a computer model (algorithm). Usually a simulation model is an abstraction that behaves somewhat like the original system, allowing thus users to replicate only a small part of the actual system under investigation (e.g. its key features or characteristics). The common perception of a simulation is that of an interactive computer program that replicates, within limits, some object, phenomenon, situation or process of the real or the imaginary world. There is a confusion between simulations and other computer applications which have a similar-looking output, like animations or visualizations. Simulations differ substantially because they predict an output based on a series of inputs. On the other hand, computer animations do not use any underlying model to calculate the behaviour of the system while they simply display a series of pre-calculated values. In conclusion, there are two key features which define a computer simulation (Thomas & Milligan, 2004):  A computer model of a real or theoretical system that contains information on how the system behaves (formal entities, properties, and rules or relationships among them).  Experimentation can take place, e.g. the user can change the input to the model affecting thus its output behaviour (Figure 1). Representation 1 Input data

Simulation model

Representation 2 … Representation n

Fig.1. The simulation process

Two methods of simulation distribution are available: CD-ROM or web-based format. A CD-ROM is suitable when the simulation material needs extensive memory and it would be too time-consuming for the user to download it from the Web. On the other hand, using a Web format makes the material immediately available from virtually anywhere in the world, independently of the computer platform used. Furthermore, material on the Web can be readily updated and easily structured through hyperlinking techniques, making thus clear the relationships between the various parts. Basically, two different technologies are used to support simulations development. Multimedia simulations, mainly, use a two dimensional (2D) representation to simulate the natural world and, as a consequence, they lack realism. Alternatively, virtual reality (VR) simulations constitute a realistic 3D, highly interactive, multimedia environment in which the user becomes a participant in a computer2


generated virtual world. The key feature of VR simulations is real-time interactivity, where the computer is able to detect user inputs and instantaneously modify the virtual world in accordance with user interactions. VR simulation environments could be explorative or immersive. The latter consist of special hardware parts including head-mounted displays, motion-sensing data gloves, eye phones and others. SIMULATION DESIGN The process of designing and building a simulation is known also as modelling process. Modelling, in general, is a way of thinking and reasoning about systems. The goal of modelling is to come up with a representation that is easy to use in describing systems in a mathematically consistent manner (Fishwick, 1995). Moreover, modelling aims at the prediction and understanding of the behaviour of the target system under a range of conditions. The creation of models is a dynamically evolving cyclic process and, in general, four main stages can be identified within it (Fishwick, 1995; Thomas, 2003): Conceptual model formation: This is an abstraction process whereby the designers define a simplified representation of the system. Conceptual modelling incorporates: a) identification of the components of the systems and the boundaries or limits of the corresponding model, b) revealing and defining rules and relationships between the various entities, and c) creation of new information in cases where there is no information available. During this stage, the components of the system and their key relationships must be established (Covert et al., 2001). Finally, the designers must decide upon the characteristics of the model, i.e. its scope (application and people to use it), its type (quantitative or qualitative), its structure (discrete, continuous or hybrid), its behaviour (discrete or continuous) and its scale. For some systems, there may be an analytic solution and a set of equations that define the model adequately, in others the computer must be used to estimate a solution. Approximations or simplifications (e.g. ignoring rotation in the simulation of a projectile’s motion) are introduced to reduce complexity, computational requirements and solution’s time. Quantitative model specification: During this phase, the appropriate mathematical equations necessary to describe and model the system must be specified. These equations must be converted to an algorithm that can be executed on a computer. The computer model (algorithm) should accomplish three main properties (goals): a) validity, it must be validated to ensure that it matches the target model and its output reflects accurately the behaviour of the real world system; b) usability, to allow the researcher or whoever will use the simulation to interpret its output and understand how it works; and c) extendibility, to allow designers or any future user to adapt model for novel uses. A simulation is much more likely to be extendible if it is written and documented by taking into account this goal. Model specification is an iterative process that is repeated until a sufficiently accurate model is obtained. There are two main ways of constructing computer models: a) programming languages (e.g. C, C++, Visual Basic, Java etc.), which can be used to construct simulation models from scratch, and b) simulation tools available for different types of modelling, e.g. dynamic systems (AgentSheets, Flexim, Model-It, ModelMaker, PowerSim, SimQuest, Stella) or discrete systems (Arena, Simul8). Simulation tools are usually domain-specific and therefore cannot be used to simulate systems of other domains in a reasonable way. Special simulation or modelling languages, like 20-sim, Dymola, Modelica etc., are superior to ordinary programming languages since they provide particular functionalities (time control, set manipulation, data visualization,

3


statistics gathering and reporting etc.) and rich model libraries. Supportive utilities, such as multimedia authoring tools or VR tools, are also used. Model execution and evaluation: In this stage the simulation is executed on a computer and its performance is compared with that of the real world system. The simulation should be able to produce the same sorts of data and input/output relationships that were initially gathered. Once the computer model is constructed, the designers conduct experiments to solve problems in order to obtain a better understanding of the system. The model is regarded as being valid as long as the simulation can accurately reproduce the real world data. A variety of factors can make a simulation invalid quite apart from the fact that the underlying equations might be inaccurate (for example, errors in input data and programming, constraints of the programming language etc.). Techniques such as sensitivity analysis and validation of accuracy are used at this stage (Thomas 2003). Model use: Once the simulation has finally been completed it may be used as an empirical research tool aiming at the understanding of the system. Thorough research will inevitably reveal flows, deficiencies or inadequacies of the model and will therefore lead to model revision or redesign. This process may ultimately conclude to the development of an entirely new model. A new modelling cycle may begin again (Figure 2), as the original system could be changed by proper interventions or refinements. Implementation Redesign

Model use

Execution Evaluation

Structure Components Scale

System

Conceptual model

Computer model

Algorithm

Fig. 2. The modelling cycle

USING SCIENTIFIC SIMULATIONS Fundamentally, the role of scientific simulations is to act as a virtual laboratory. During the last decades, they have dynamically evolved as a new form of scientific tool. Regarding their applications, simulations tend to play one of the following subsidiary roles within the framework of scientific research: Complementary to empirical investigation: This can be done through experimenting or observing the natural world. There are numerous complementary benefits offered by simulations, especially in the cases of complex systems described by many interrelated variables, when the relationships between variables are nonlinear or when the underlying model contains random variables etc. Once a simulation has been validated against real-world or experimental data it can be used to pose specific questions, to predict the outcomes of real experiments, and provide a quantitative basis of system study in order to help researchers in achieving qualitative

4


understanding (Covert et al., 2001; Hanan et al., 2002; Washington et al., 2000; Wiechert, 2002). Superior to empirical investigation: There are numerous reasons as to why simulations are considered by many researchers as superior to the experimental or direct observation of a system under study (Meza et al, 2003; Washington et al.,); namely, simulations a) allow more control, b) allow more detailed investigation, c) offer time compression or expansion, and d) constitute a less dangerous and costly investigation tool. The only method for empirical investigation: There are various reasons that do not allow an experiment or an observation to take place in order to investigate a system (Feinstein & Parks, 2002; Axelrod, 1997). For example simulations offer the only possibilities for experimentation on ecosystems (Grant & Thompson, 1997; Lorek & Sonnenschein, 1999). There are also ethical or safety concerns, e.g. medical training (Gibson et al., 1998), car crash accident (Simpson et al., 2003) or a nuclear war study. Finally, in many cases there are needs for time expansion or compression (a fast fire starting or the development of a new forest), spatial scale problems or extreme conditions not easily reached in laboratory settings, e.g. galaxy formation (Meza et al., 2003), climate models (Washington et al., 2000) etc. EDUCATIONAL SIMULATIONS Scientists create simulations on various domains, such as technology, physics, chemistry, biology, medicine, environmental sciences, social and financial studies etc. to facilitate their research, to express and test their theories, and to improve their understanding about complex systems. Scientific simulations are, usually, too complex to be used in the various educational contexts. Educational simulations are created in order to facilitate students’ or trainees’ learning. Because of its purpose, an educational simulation is an abstracted representation of the target system, which is neither as complex nor as realistic as the corresponding scientific one. De Jong & Joolingen (1998) have divided educational simulations in two main categories: a) Operational simulations Operational simulations are designed to facilitate the construction of practical knowledge, for example in areas such as medical training, pilot training etc. (Gibson et al., 1998; Leemkuil et al., 2000). They are based on operational models which allow students to both practically and psychologically play the role for which they were trained (e.g. the role of a surgeon or a pilot). Operational simulations often use non-standard input and output mechanisms. b) Conceptual simulations On the other hand, conceptual simulations are designed to facilitate conceptual knowledge construction on the part of the students. They are based on conceptual models (used within subject domain education), which simulate the relationships that exist between the variables of a real-world system while, at the same time, allow the user (student) to manipulate those variables. There are four main building blocks in an educational simulation (Alessi, 1988): the underlying model, the various presentations, the user actions and the system feedback (human-computer interaction). According to Maier & Grobler (2000) computer simulations have three key aspects: the underlying mathematical model, the human–computer interface and the various functionalities. They consider underlying model as being a major part of a computer simulation, and they place the rest of

5


Alessi’s components under a new category which comprises the characteristics of a human-computer interface. They have also added a new component, named functionality, in order to cover those characteristics of computer simulations that are not determined by the underlying model or the human-computer interface. For example, functionality involves access to additional source materials, the extent to which the structure of the underlying model is explicitly shown (degrees of transparency), the progress of time within the simulation (time-step) etc. Conceptual simulations are considered to be effective educational environments aiming at enhancing teachers’ instructional potentialities and provoking students’ active engagement. They have been proposed as effective tools for supporting students’ understanding since they behave in a similar way to the system or the process modelled. They offer a great variety of opportunities for modelling concepts and processes and therefore provide a bridge between students’ prior knowledge and the learning of new physical concepts, and help students develop scientific understanding through an active reformulation of their misconceptions (de Jong and Joolingen, 1998; Jonassen et al., 2003; Jimoyiannis & Komis, 2001). Designing efficient educational simulations In the attempt to describe the building blocks of an educational simulation, within the framework described by Alessi (1988), three main components can be identified (Figure 3): the simulation scenario, the mathematical model of the target system and the instructional overlay. The components above collectively define the body of knowledge to be learned and how learning is approached or supported by the simulation environment. Scenario Educational Simulation

Model Instructional overlay

Fig. 3. The basic components of an educational simulation

Scenario related issues concern model focus and the degree of model modification and model transparency incorporated (for teachers and students also). In many cases, teachers prefer to modify a modelling scenario in order to provide a different task to their students, stimulate their involvement, make them being more focused on a particular aspect of the model, describe different phenomena, etc. It is preferable that students at different educational levels, and students who have different skills, interests or educational needs could use the same model. For example, the teacher may wish a model that could be simplified for novice users, allowing them to access a limited number of model variables, or to vary the degree to which students are allowed to view or manipulate the underlying mathematical model. Various simulation tools, like Interactive Physics, Modellus, Tina, ChemLab, Modelling Space, SimForest, Explore-It ect., among others, offer those possibilities as well. The mathematical models, upon which operational simulations are based, need to be as realistic as possible since they are being used for training in specific real-world

6


procedures such as medical procedures (Gibson et al., 1998). On the other hand, the mathematical models upon which conceptual simulations are based are usually simplified to some degree in order to facilitate learning. Such a simplification is necessary because by increasing complexity increases the time required by students to understand the simulation, and so does the likelihood that they will become frustrated and demotivated (Bos, 2001). By reducing simulation realism, therefore, the conceptual simulation designer can clarify complex or difficult concepts and can tailor the simulation to the students’ prior knowledge and experience. The third component of a conceptual simulation, the instructional overlay, is made up of those features defining the educational context and the representational forms used, the educational approach (discovery or expository learning) and tasks used, and, finally, learner motivation, guidance, assistance and feedback (Hmelo & Day, 1999; Leemkuil et al., 2000; Lee et al., 2004). A well-designed instructional overlay can: a) prompt and motivate students, b) incorporate questions that will direct students towards educational goals, c) focus students’ attention upon educationally important aspects of the simulation, and d) progressively unfold the complexity of a simulation over a series of stages in order that students not be overwhelmed by it. Creating a model of a single object or event is, usually, an easy task. However, a simulation model of an entire virtual world, in which all objects and events are represented in a consistent and complimentary manner, requires a great deal of mental and creative effort. Moreover, developing computer-based simulations for learning is not only a technical matter; however, such systems should combine both modelling and instructional knowledge with a pedagogical strategy (Joolingen & de Jong, 2003). The following interrelated parameters (phases) could define a consistent framework for building efficient education simulations: Instructional design: First of all, one has to estimate the learners’ attributes and their prior subject knowledge, and also to define the learning goals associated to the body of domain knowledge to be learned. Following is the design of students’ activities that will be performed, and the estimation of the didactical knowledge or any other relevant information from the Didactics of the subject matter. Learner interface: During this phase we create the simulation interface and embody the tools appropriate to visualize and represent knowledge effectively for the learners (graphs, tables, vector or textual representations). An educational simulation must offer different views of the underlying model, from textual to graphical representations. As Ainsworth (1999) has shown, different representations used simultaneously can constrain interpretation, construct deeper understanding or complement each other. Pedagogical strategies: In general, pedagogical strategies define how learning is approached within the system (for example exploratory or discovery learning; this determines the models to be discovered or experienced by the learners). Another issue is the instructional interventions to be incorporated (e.g. explanations, exercises, tasks or other type interventions). Simulation model creation: This phase incorporates identification and definition of the underlying simulation model (parameters, relationships among them, limits, representations used, models’ type, structure and behaviour, user interface interactions, student supporting tools, etc.) Integration: All the information and the outcomes above should be integrated to a complete system capable to support students’ meaningful learning and teachers’ role as efficient facilitators (Fig. 4).

7


Learning goals

Knowledge body

Learner attributes

Simulation attributes

Pedagogical strategy

Integration Figure 4. Educational simulation aspects

CONSTRUCTIVIST LEARNING BY USING SIMULATIONS Over the last decades the constructivist view of learning is dominant among scholars and researchers (Driver, 1989; Duit & Treagust, 2003; Von Glasersfeld, 1987). According to this view, students do not passively absorb information but meaningful learning, rather, occurs through an active process of creation and modification of knowledge structures. Students achieve learning through using their existing knowledge, beliefs, interests and goals, in order to interpret any new information. This may result in modifying or revising their ideas. In this way, learning occurs as each individual’s conceptual schemes are progressively reconstructed while he/she becomes exposed to new experiences and ideas. There are two main constructivist schools: a) Cognitive constructivism which places emphasis on the personal construction of knowledge. According to this view, teachers have a relatively peripheral role in providing suitable experiences that will facilitate learning; b) on the other hand, according to social constructivism, knowledge is socially constructed and learning takes place in particular social and cultural contexts (Vygotsky, 1978). In both cases, the emphasis is on interactive-rich educational environments where students are given opportunities to interact with adults, peers and knowledge, in order to negotiate their meaning. According to this view, teachers have a central role in providing guidance and support to learners (scaffolding). Through this process a teacher can gradually guide his/her students to develop their knowledge and skills while making connections with their pre-existing mental models and schemes. Within a constructivist framework of learning, simulations may offer strong benefits not only by facilitating constructivist-learning activities but also by supporting different types of learners. Many researchers have advocated the educational potential of computer-based simulations, based on the fact that the latter provide opportunities for active learning (de Jong & Joolingen, 1998; Jonassen et al., 2003), enable students to perform at higher cognitive levels (Huppert et al., 1998) and promote conceptual change (Tao & Gunstone, 1999; Jimoyiannis & Komis, 2001). For instance, students may vary a selection of input parameters and observe the extent to which each individual parameter affects the whole system. Alternatively, they can explore combinations of parameters and observe their effect on the evolvement of the system. Educational computer simulations constitute open learning environments that provide students with opportunities to (Jimoyiannis & Komis, 2001):

8


 access, visualize and investigate dynamic phenomena and situations which are, otherwise, difficult to experience in a classroom or lab setting, because they are extremely complex, technically difficult, dangerous, unethical, money- or timeconsuming, happen too fast or very slowly etc.  introduce simplified concepts prior to developing complexity  support a qualitative understanding of complex systems knowledge  develop a conceptual understanding about phenomena and laws through an active process of hypothesis-making, and ideas testing, by changing the input variables and observing the effect directly on the output  isolate and manipulate parameters in order to develop meaningful understanding of the relationships between concepts, variables and phenomena  employ a variety of representations (pictures, animation, graphs, vectors and numerical data displays) which are efficient in understanding the underlying concepts, relations and processes  express their own representations and mental models about real world phenomena, situations or systems. The constructivist perspective of learning argues that knowledge is not transferred to students’ minds but it is rather achieved by constructing efficient models of the natural world. Hestenes (1992) has distinguished two types of models: a) mental models, which are representations of the physical phenomena constructed by students containing a set of information about what they know (either correct or incorrect) and b) conceptual models, which originate from mental models and are created by the cooperative activities of scientists and domain specialists. These are objective representations in the sense that they are independent of any particular individual. There is also a third type of models, constructed by domain and education specialists, one could name as didactical models since they aim at helping students to achieve conceptual understanding. Students’ active engagement with didactical models is essential to overcome conceptual obstacles and construct scientifically valid conceptual models. Bliss (1996) has also analyzed the differences between two types of models in educational simulations. a) Exploratory models: They are constructed by experts to represent domain knowledge and/or simulate complex processes and laws. These environments encourage students to explore and interact with them, to handle various input parameters and, finally, to observe their results. b) Expressive models: They allow students to express their own ideas on a domain. They provide learners with tools to define relationships between concepts, to explore the consequences of those student-defined relationships, and learn through an active process of representing their own models. In this framework we have suggested that educational simulations can be used in three different ways (Jimoyiannis & Komis, 2001):  as an artifact, which helps teachers to demonstrate explicit visual representations of complex systems and relations, in order to support their lecture or explanations  as a virtual laboratory representing an exploratory world where students can use models to conduct experimentation, to create and test hypotheses and, finally, construct their own understanding  as a student activity, which aims at students’ constructing and presenting their own (expressive) models through a meaningful learning experience.

9


Computer simulations have been successfully applied from primary and secondary school (Davies, 2002; Huppert et al., 1998; Jimoyiannis et al., 2000; Pallant, 2004; Tao, 1997) to university education (Schroeder, 1993; Granlund et al., 2000; Hundhausen et al., 2002; Warner et al., 2000) covering various disciplines. There are many studies providing us with supportive evidence regarding the effect of computer simulations to activate students, to motivate them in order to perform at higher cognitive levels (creative and analytic thinking, skills development, problem solving skills and abilities etc.) and to promote conceptual change and knowledge construction. Bakas & Mikropoulos (2003) have studied the effect of a VR simulation environment on students’ comprehension of planetary phenomena. In their study Peña & Alessi (1999) have showed that simulations were equally effective to microcomputer based labs in facilitating comprehension of the concepts involved in objects’ free fall. The combination of simulations and the Web can create powerful and dynamic learning environments (Lee et al., 2004). Efficient Java applets may be found on many educational sites and are frequently used as supplements to classroom lectures and traditional labs (e.g. JeLSIM, Physlets, Easy Java Simulations etc.).

Figure 5. Simulation of a satellite moving around the earth through Interactive Physics

Figure 5 shows a screenshot of Interactive Physics presenting the simulation of a satellite moving around the earth. This simulation is a model paradigm that can be used according to the ways shown above, e.g. as a visual representation to support teacher’s instruction, as a virtual laboratory for students to explore and study this phenomenon, and as a student activity to build it from scratch. This last activity gives the students the opportunity to think scientifically about the behavior of complex systems, to reflect upon their own understanding, to test and refine their own mental models, and, finally, to construct valid conceptual models. Simulations of this type

10


have been used to support secondary school students’ learning in the domain of kinematics (Jimoyiannis & Komis, 2001). Students in the experimental group have been engaged in tasks which demand the exploration of simulations through Interactive Physics. Results exhibited that students significantly improved their achievement rates in tasks regarding the concepts of velocity and acceleration. Computer simulations can effectively support students and novice programmers to achieve meaningful understanding of the basic algorithmic concepts and develop their programming skills (Hundhausen et al., 2002; Sajaniemi & Kuittinen, 2005, Jimoyiannis et al., 2006). PlanAni is an algorithm visualization environment based on the idea of teaching students the roles of variables rather than the programs as a single unit (Sajaniemi, 2002). This notion is based on the fact that variables are not used in a random or ad-hoc way but there are several standard use patterns that occur over and over again in the various algorithms.

Figure 6. Simulation of the bubble-sort algorithm through PlanAni (Gr)

PlanAni(GR) is a new version of this simulation environment adapted to the Greek secondary school needs (Jimoyiannis et al., 2006). In Greek upper secondary schools an ideal programming language, named GLOSSA, is used to support introductory programming lessons and students in developing their own algorithms. Figure 6 shows a screenshot of the PlanAni(GR) interface animating the execution of the bubble-sort algorithm. The software provides students with opportunities to:  visualize and simulate the overall execution of the program and the dynamic change of the values of the variables involved  input data and observe on the screen the way in which the execution output is produced  achieve different representations of the variables according to their roles in the program  isolate and visualize statements or complex structures, such as conditional structures or iterations (loops)

11


 receive explanatory messages of the effect and the role that each variable or statement has in the program  use the tools embodied (start, stop, simulation speed, stepwise execution of the program) in order to achieve meaningful understanding. Despite all these advantages and possibilities that simulations offer, teachers seem not to use them extensively in their instruction (Jimoyiannis & Komis, 2006) while they encounter many difficulties to support simulation-based discovery learning and scaffold students’ activities. De Jong and Joolingen (1998) have provided an extensive overview of the main problems and indicated how simulations could be extended to overcome these problems. Their reasoning is anchored on the basis of integrating discovery learning with instructional support. I further argue that, despite their potential, simulation-based learning environments cannot guarantee effective learning without sufficient support, both for students and teachers. However, constructing a dynamic model is a complex task and it is reasonable that students, as novice modellers, encounter many difficulties in accomplishing this. Sins et al. (2005) have studied students who work in dyads, on a modelling task in the domain of physics. Their results indicate that the successful students differed from the less successful ones in using more prior knowledge and in showing more inductive reasoning. They have pointed out the importance of students’ support and suggested that efficient scaffolding should: a) encourage students to activate their prior knowledge before and during their modelling activities, and b) motivate deep reasoning about their models by testing them against multiple datasets. Students should be asked to model phenomena of which they already have knowledge or experience. Zang et al. (2004) proposed a triple learning support scheme, for scientific discovery learning based on computer simulations which involves: a) interpretative support that helps learners with knowledge access and activation, the generation of appropriate hypotheses, and the construction of meaningful and coherent understanding, b) experimental support that scaffolds learners in the systematic and logical design of scientific experiments, the prediction and observation of outcomes, and the drawing of reasonable conclusions, c) reflective support that increases selfawareness of the discovery processes, and prompts their reflective abstraction and knowledge integration. CONCLUSION In this article simulations’ exploration and modelling have been considered as a scientific activity, which can help students make their mental models explicit, develop creative and flexible thinking, and achieve meaningful learning. This is not an ideal activity while students and teachers encounter many difficulties. There is still a lot to learn about simulation-based learning, for example:  What type of conceptual or other difficulties do students encounter? In what ways students’ scaffolding should be addressed?  What prevents teachers from using simulation and modelling activities in their classes?  What type of support do teachers need? Is only their technological support needed or are there important pedagogical issues that must be equivalently analyzed?  What types of interventions are needed in the curriculum? Many researchers assert that ICT integration in the educational practice is a complex and multi-faceted issue and teachers constitute a critical factor (Kumar & 12


Kumar, 2003; Russell et al., 2003; Jimoyiannis & Komis, 2006a). Teachers must be able, not only to use ICT tools, but also to principally reorganize their instruction by using student-centered activities, based on appropriate ICT applications. Their preparation and support programs, must clearly articulate specific types of effective instructional models and representative paradigms of ICT use, for every subjectmatter in the curriculum. It has been shown that the functionality of computers in the class has been quite different for teachers of different attributes such as gender, age, subject specialty, computer use and teaching experience (Jimoyiannis & Komis, 2006a; 2006b). Professional and pre-service development programs should focus on changing teachers’ pedagogical cultures and philosophies about the teaching and learning processes through training them on how to use appropriate ICT tools with their students. This must be organized in a framework of broader instructional reforms aiming at the curriculum, the educational media and, principally, at the pedagogical practices used. I therefore argue that it is important to consider the use of simulation-based learning environments carefully and that there is a sound justification for the pedagogical strategies used. Clearly there is no single technology or instructional approach that can resolve all problems and meet all pedagogical needs. To increase the likelihood that simulations will be used in the school practice effectively, teachers need to be encouraged to try and acquire positive experiences about their effectiveness on teaching and learning. REFERENCES Ainsworth, S. (1999). The functions of multiple representations. Computers and Education, 33, 131152. Alessi, S. M. (1998). Fidelity in the design of instructional simulations. Journal of Computer-Based Instruction, 15(2), 40-47. Axelrod, R. (1997). Advancing the art of simulation in the social sciences. Obtaining, analyzing, and sharing results of computer models. Complexity, 3(2), 16-22. Bakas, C., & Mikropoulos, T. A. (2003). Design of virtual environments for the comprehension of planetary phenomena based on students’ ideas. International Journal of Science Education. 25(8), 949–967. Bliss, J. (1996). Externalizing thinking through modeling: ESRC tools for exploratory learning research program, in S. Vosniadou, E. de Corte, R. Glaser & H. Mandl (eds.), International perspectives on the design of technology-supported learning environments, 25-40, New Jersey: Lawrence Erlbaum Associates. Bos N. (2001). What do game designers know about scaffolding? Borrowing SimCity design principles for education. Technical report written for the PlaySpace working group. Center for Innovative Learning Technologies, Retrieved January 12, 2006, http://wwwpersonal.si.umich.edu/~serp/work/SimCity.pdf Davies, C. H. J. (2002). Student engagement with simulations: a case study, Computers and Education, 39, 271-282. de Jong, T., & Joolingen, W. R. (1998). Scientific discovery learning with computer simulations of conceptual domains, Review of Educational Research, 68(2), 179-202. Driver, R. (1989). Students’ conceptions and the learning of science. International Journal of Science Education, 11, 481-490. Duit, R., & Treagust, D. F. (2003). Conceptual change: A powerful framework for improving science teaching and learning. International Journal of Science Education, 25, 671-688. Feinstein, A. H., & Park S. J. (2002). The use of simulation in hospitality as an analytic tool and instructional system: a review of the literature, Journal of Hospitality & Tourism Research, 26(4), 396-421. Fishwick, P. A. (1995). Simulation model design and execution. Prentice Hall.

13


Granlund, R., Berglund, E., & Eriksson, H. (2000). Designing web-based simulation for learning, Future Generation Computer Systems, 17, 171-185. Grant, W. E. & Thompson, P. B. (1997). Integrated ecological models: Simulation of socio-cultural constraints on ecological dynamics. Ecological Modelling, 100, 43-59. Gibson, S., Fyock, C., Grimson, E., Kanade, T., Kikinis, R., Lauer, H., McKenzie, N., Mor, A., Nakajima, S., Ohkami, H., Osborne, R., Samosky, J., & Sawada, A. (1998). Volumetic object modelling for surgical simulation. Medical Image Analysis, 2(2), 121-132. Hanan, J., Prusinkiewicz P., Zalucki M., & Skirvin, D. (2002). Simulation of insect movement with respect to plant architecture and morphogenesis. Computers and Electronics in Agriculture, 35, 255-269. Hestenes, D., (1992). Modeling games in the Newtonian World. American Journal of Physics, 60, 732748. Hmelo, C., & Day, R. (1999). Contextualized questioning to scaffold learning from simulations. Computers and Education, 32, 151-164. Hundhausen, C. D, Douglas S. A., & Stasko, J. T. (2002). A meta-study of algorithm visualization effectiveness, Journal of Visual Languages and Computing, 13, 259-290. Huppert, J., Yaakobi, J. & Lazarowitz, R. (1998). Learning microbiology with computer simulations: students’ academic achievement by method and gender, Research in Science and Technological Education, 16, 231-245. Jimoyiannis, A., & Komis, V. (2001). Computer simulations in physics teaching and learning: A case study on student’s understanding of trajectory motion. Computers and Education, 36, 183-204. Jimoyiannis, A., & Komis, V. (2006a). Exploring secondary education teachers’ attitudes and beliefs towards ICT in education, THEMES in Education, 7(2), 181-204 Jimoyiannis A. and Komis V. (2006b). Factors affecting teachers’ views and perceptions of ICT in education. In P. Isaias, M. McPherson & F. Banister (Eds.), Proceedings of the IADIS International Conference e-Society 2006, Vol. I, 136-143, Dublin. Jimoyiannis, A., Mikropoulos, T. A. & Ravanis, K. (2000). Students’ performance towards computer simulations on kinematics. Themes in Education, 1(4), 357-372. Jimoyiannis, A., Tsiotakis, P. & Sajaniemi, J. (2006). Investigating the role of algorithm simulations in teaching programming at secondary education level, Pan-Hellenic Conference ‘’Digital Educational Media: Development, Application and Evaluation Issues’’, Volos, Greece (in Greek). Jonassen D. H., Howland J., Moore J., &. Marra, R. M (2003). Learning to solve problems with technology: A constructivist perspective, Prentice Hall. Joolingen, W. R. & de Jong, T. (2003). SimQuest, authoring educational simulations, In T. Murray, S. Blessing, S. Ainsworth (Eds.), Authoring Tools for Advanced Technology Learning Environments: Toward cost-effective adaptive, interactive, and intelligent educational software, 1-31, Dordrecht: Kluwer. Kumar, P. & Kumar, A. (2003). Effect of a Web-base project on preservice and inservise teacher’ attitude toward computers and their technology skills. Journal of Computing in Teacher Education, 19(3), 87-92. Lee, K. M., Nocoll, G. & Brooks, D. W. (2004). A comparison of inquiry and worked example Webbased instruction using Physlets. Journal of Science Education and Technology, 13(1), 81-88. Leemkuil, H., de Jong, T., & Ootes, S. (2000). Review of educational use of games and simulations, Retrieved January 12, 2006, http://kits.edte.utwente.nl/documents/D1.pdf Lorek, H., & Sonnenschein, M. (1999). Modelling and simulation software to support individual-based ecological modelling. Ecological Modelling, 115, 199-216. Maier F. H. & Grobler A. (2000). What are we talking about ? A taxonomy of computer simulations to support learning, System Dynamics Review, 16(2), 135-148. Meza, A., Navarro, J. F., Steinmetz, M., & Eke, V. R. (2003). Simulations of galaxy formation in a ΛCMD universe. The dissipative formation of the elliptical galaxy. The Astrophysical Journal, 590, 619-635. Merians, A. S., Jack, D., Boian R., Tremaine M., Burdea G. C., Adamovich, S. V., Recce M., & Poizner H. (2002). Virtual reality-augmented rehabilitation for patients following stroke. Physical Therapy, 82(9), 898-915. Pallant, A. & Tinker R. F. (2004). Reasoning with atomic-scale molecular dynamic models. Journal of Science Education and Technology, 13(1), 51-66. Peña, C. M. & Alessi, S. M. (1999). Promoting a qualitative understanding of Physics. Journal of Computers in Mathematics and Science Teaching, 18 (4), 439-457.

14


Russell M., Bebell D., O’Dwyer L. & O’Connor, K. (2003). Examining teacher technology use. Implications for preservice and inservice teacher preparation. Journal of Teacher Education, 54(4), 297-310. Sajaniemi, J. (2002). An empirical analysis of roles of variables in novice-level procedural programs, In Proceedings of IEEE 2002 Symposia on Human Centric Computing Languages and Environments, 37-39, Arlington, Virginia, USA. Sajaniemi, J. & Kuittinen, M. (2005). An experiment on using roles of variables in teaching introductory programming, Computer Science Education, 15, 59-82. Schroeder, D. V. & Moore, T. A., (1993). A computer-simulated Stern-Gerlach laboratory. American Journal of Physics 61, 798-805. Simpson, G., Johnston L., & Richardson M. (2003). Tan investigation of road crossing in a virtual environment. Accident Analysis and Prevention, 35, 787-796. Sins, P. H. M., Savelsbergh, E. R., & van Joolingen, W. R. (2005). The difficult process of scientific modelling: An analysis of novices’ reasoning during computer-based modelling. International Journal of Science Education, 27(14), 1695-1721. Tao, P. K. (1997). Confronting students’ alternative conceptions in mechanics with the Force and Motion Microworld. Computers in Physics, 11(2), 199-207. Tao, P. K. & Gunstone, R. F. (1999). Conceptual change in science through collaborative learning at the computer, International Journal of Science Education, 21, 39-57. Teodoro, V. D. (1993). Learning with computer-based exploratory environments in Science and Mathematics. in S. Vosniadou, E. de Corte & H. Mandl (eds.), Technology-based learning environments, NATO ASI Series F, 137, 26-32, Berlin: Springer-Verlag. Thomas R. (2003). What are simulations, Retrieved January 25, 2006. http://www.jelsim.org/resources/whataresimulations.pdf Thomas R. C. & Milligan C. D. (2004). Putting teachers in the loop: tools for creating and customizing simulations, Journal of Interactive Media in Education, 15, http://www-jime.open.ac.uk/2004/15 Warner S., Catterall, S., Gregor, E., & Lipson E. (2000). SimScience: Interactive educational modules based on large simulations. Computer Physics Communications, 127, 1-5. Washington, W. M., Weatherly, J. W., Meehl, G. A. Semtner, A. J., Bettge, T. W., Craig, A. P., Strand, W. G., Arblaster, J., Wayland, V. B., James. R., & Zhang, Y. (2000). Parallel climate model (PCM) control and transient simulations. Climate Dynamics. 16, 755-774. Wiechert, W. (2002). Modeling and simulation: tools for metabolic engineering. Journal of Biotechnology. 94, 37-63. Von Glasersfeld, E. (1987). Learning as a constructive activity. In C. Janvier (Ed.), Problems of representation in the teaching and learning of mathematics. 3-17, Hillsdale, NJ: Erlbaum. Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press. Zhang, J., Chen, Q., Sun, Y., & Reid D. J. (2004). Triple scheme of learning support design for scientific discovery learning based on computer simulation: experimental research. Journal of Computer Assisted Learning. 20, 269-282. Ziv, A., Small, S. D., & Wolpe, P. R. (2000). Patient safety and simulation-based medical education. Medical Teacher. 22(5), 489-495.

15


APPENDIX. Table 1. Simulation tools and environments mentioned in the text Simulation-modelling environment 20-sim AgentSheets Arena ChemLab Dymola Easy Java Simulations eM_Plant Explore-It Flexsim Interactive Physics JeLSIM Modelica Model-It ModellingSpace Modellus ModelMaker Physlets PowerSim SimForest SimQuest Simul8 Simulink Stella TINA Vensim

URL http://www.20sim.com http://www.agentsheets.com http://www.arenasimulation.com http://modelscience.com http://www.dynasim.se http://fem.um.es/Ejs/Ejs_en/index.html http://www.emplant.de/simulation.html http://exploreit.com/aboutUs/overview.htm http://flexsim.com http://www.krev.com http://www.jelsim.org/ http://www.modelica.org http://hi-ce.org/modelit http://www.modellingspace.net http://phoenix.sce.fct.unl.pt/modellus http://www.modelmakertools.com/ http://webphysics.davidson.edu/Applets/Applets.html http://www.powersim.com http://ddc.hampshire.edu/simforest/software/software.html http://www.simquest.com http://www.simul8.com/ http://www.mathworks.com/products/simulink http://www.iseesystems.com/softwares/Education/StellaSoftware.aspx http://www.tina.com http://www.vensim.com/software.html

16


KEY TERMS Scientific simulation: It is a technique of imitating the behaviour of an actual or theoretical system by means of an analogous mathematical model. In the simplest sense, a system is a set of interacting identities. The mathematical equations that produce the model represent the various processes taking place within the target system. Currently simulation uses cover a wide range of applications within the areas of research, analysis studies, system design, training and education, entertainment etc. Computer simulation: In a computer simulation proper executable algorithms produce the underlying mathematical model. There are two key features defining a computer simulation: a) a computer model of the target system that contains information on how the system behaves (formal entities, properties, rules and relationships among them), and b) experimentation can take place, e.g. the user can change the input to the model affecting thus its output behaviour. Educational simulation: It is a computer simulation created to facilitate learning on the part of students or trainees. Educational simulations are abstracted or simplified representations of a target system, which are neither as complex nor as realistic as the relevant scientific simulations. Operational simulations: They are educational simulations designed to facilitate the construction of practical knowledge, for example in areas such as medical training, pilot training etc. They are based on operational models, which use non-standard input and output mechanisms. Conceptual simulations: Conceptual simulations are educational simulations designed to facilitate conceptual knowledge construction on the part of the students. They are based on conceptual models which: a) simulate the relationships that exist between the variables of a real-world system and b) allow the students to manipulate those variables. There are three main components in a conceptual simulation: the simulation scenario, the mathematical model of the target system and the instructional overlay. Instructional overlay: It is the component of a conceptual simulation, characterized by those features defining the educational context, the representational forms, the educational approach and the tasks used. A well-designed instructional overlay should: a) prompt and motivate students’ engagement, b) incorporate feedback and other tools that may assist and guide students towards the educational goals, c) focus students’ attention upon cognitively important aspects of the simulation, and d) unfold the complexity of the simulation over a series of stages in order that students not be overloaded or overwhelmed. Mental models: They are schemata and representations of the real world phenomena constructed by students. They contain a set of information about what students already know, either correct or incorrect. Students usually exhibit mental models, which are not consistent with the relevant scientific paradigms. Conceptual models: They are mental models created by the cooperative activities of scientists and domain specialists. These are objective representations in the sense that they are consistent with the relevant scientific paradigms.

17


Virtual reality (VR): It is a three dimensional (3D), realistic and highly interactive multimedia environment, in which the user becomes a participant in a computergenerated virtual world. The key feature of a VR simulation is its real-time interactivity, where the computer is able to detect user inputs and instantaneously modify the virtual world in accordance to user interactions. VR-based simulation environments could be explorative or immersive (which consist of special hardware parts including head-mounted displays, motion-sensing data gloves, eye phones etc). Constructivism: According to the constructivist view of learning, students do not passively absorb information but, rather, meaningful learning occurs through an active construction and modification of their knowledge structures. When students are learning they use their existing knowledge, beliefs, interests, and goals to interpret any new information, and this may result in their ideas becoming modified or revised. There are two main constructivist schools: a) cognitive constructivism, which emphasizes on the personal construction of knowledge; b) social constructivism, which emphasizes on knowledge construction in particular social and cultural contexts. In both cases, the emphasis is on interactive environments where students are given opportunities to negotiate their ideas and meanings. According to this view, teachers have a central role in providing guidance and support to their students (scaffolding).

18