Summary

ModellingSpace was a project partially funded by the European Comission. The consortium was formed by five Universities and one private company: University of Aegean (Greece); University of Patras (Greece); University of Mons-Hainaut (Belgium); University of Lisbon (Portugal); University of Angers (France); and SchlumbergerSema (Spain) - now Atos Origin.

ModellingSpace objective was to develop an open learning environment which permits modelling activities by young students, supporting them to express their ideas, to design and test models of various situations, and allowing them to collaborate with other students and teachers via internet.

The work was based on a previous project, Modelscreator, carried out in Greece and partially funded by the Greek government.

Short Description

Epistemological and learning order reasons suggest the design and development of an open learning environment which permits modelling activities by young students, supporting them to express their ideas, to design and test models of various situations, while they could collaborate with other students and teachers via Internet. This is the objective of the ModellingSpace project.

ModellingSpace objectives

ModellingSpace constitutes a complete open learning system, adaptable to a wide range of students (11-17 years old), able to be used during different curriculum subjects, school classes and European countries.

Its purpose is to support students as well as teachers during learning/teaching activities, and to permit the modelling of situations studied in mathematics, physics, chemistry, biology, Environmental education, which are related to the existing national curricula and other interdisciplinary situations.

The system & the accompanied material have been delivered in 4 European languages (English, French, Portuguese, Greek).

ModelsCreator

Modelscreator is a modelling software developed in Greek and is already successfully tested in many Greek schools. It has been the previous step to the development of the ModellingSpace.

Modelscreator offers the students a number of situations which are proposed for modelling, and permits the design, testing and validation of models.

A student who wants to design a model must first determine the model's entities, located at the left of the screen, and the list of relations, which are located on the right side of the screen.

For the building of models Modelscreator offers concrete entities and abstract ones (which represent objects and concepts); and four categories of different relations to establish among the entities:
a) qualitative logical relations (expressed by logical operators)
b) qualitative semantic relations (able to produce concept maps).
c) semi-quantitative relations (relations in terms of variation of proper
d) quantitative simple algebraic relations (defined by arithmetic operators).

Once the students have chosen the entities and the relations which represent the phenomenon he wants to simulate, Modelscreator offers them the chance to run a dynamic model. While it runs, a simulation of the modelled phenomenon appears in the area of entities' icons. The examples mentioned in Dimitracopoulou et al;AI-ED 99, S.P Lajoie & M.Vivet (Eds), pp.109-121 and illustrates this:

1) In the first example, shown in the previous figure, when the model runs, the student can see the water filling up the barrel for as long as the tap is turned on. In the relation between the barrel's volume and tap's rate of flow is an inverse analogy, it will result in the decrease of the water's volume in the barrel while the tap is on.

2) The second example (see next figure) is a case of decision models. The student can see- in the icon related to the effect ('then')- the simulation of the decision's consequences (a boy will either cross the street safely or will have an accident.

For more information about Modelscreator go to Modelscreator Website.

Innovation

The innovation introduced by the proposed project can be distinguished to its aspects concerning: the modelling system (its components, characteristics and functionality), its web-based collaborative space, the multinational research educational application and its technological aspects interleaved with all the above.

i) Innovative aspects concerning modelling system to be developed (for 11-17 years old students) The existing modelling software can be classified into three wide categories, which support the corresponding reasoning modes: quantitative (these models work on countable things and algebraic forms reflecting the connections between them), qualitative (these models represent the knowledge, that is not possible to be reflected in a countable way and involves usually categorical distinctions and decision-making) and semi-quantitative modelling. The latter class of models, even if it depends on countable objects, it does not reflect their values. Scientists from science education and psychology fields have conceived semi-quantitative modelling, offering an intermediary tool for the children and helping them to have progressively access to the quantitative reasoning.

During the last decade, the interest in modelling activities and the possibilities offered by technology have led to the development of a number of systems that concern different kinds of modelling. All of these systems usually support only one type of modelling (dynamic, space distribution, logic, or qualitative), only some of them focus on special domains and none of them supports by distance teaching and learning.

Our endeavour to design and develop the MODELLINGSPACE is due to some main considerations/design principles, which have not -so far- been accomplished by other existing systems:

Concerning modelling, two main principles are specified:

• Expression through qualitative, semi-quantitative and quantitative reasoning.
• Incorporation of different categories of models (semantic models as concept maps and logic formalism, semi-quantitative modelling, algebraic formalism) in a simplified and synthetic mode.

Concerning support on young students reasoning expression and evolution in different ages, the following principles are determined:

• The expression through the greatest visualisation: the application of this principle concerns the entities, their properties and the properties' relations.
• The combination of modelling tools with real world simulations, not abstract ones: The simulations (necessary in order to validate some models) that are being produced from most of the existing modelling systems are merely abstract.
• The incorporation of alternative and multiple forms of representations: the alternative forms of representation concern the models (with their entities) as well as the different kinds of data produced by models. The multiple representations provide cognitive assistance for reasoning and consequently for learning.
• The support of the development of metacognitive skills: the provided tools a) promote the metaconceptual awareness by inviting students to write down their thoughts and especially their predictions and interpretations on a special formed notebook, b) offer possibilities of meta-analysis.

Concerning interface design and human - computer interaction for young students, the following requirements are specified:

• Support of the distinction of different actions and functions during modelling process by appropriate distinction of working areas and types of models.
• Direct manipulation interface and ergonomics suitable for young students.
• Multi-help-system: taking into account that children need to have a direct help, not always in text mode.

Concerning teachers support and general educational management

• Tutoring or coaching when students interact or collaborate via a technology-based learning environment, it constitutes a merely complex task. It is needed to really support teachers by appropriate tools such as: "Student's activities historic", "the automatic comparison of students work" a "New entities Editor", etc.

The MODELLINGSPACE architecture will consist of some main components and underlined tools:

1. The component "Study Themes Creation Tool" will containa number of situations that are proposed for modelling, while it will allow for the creation of new problems by teachers and students (using a special multimedia editor).
2. The "Models Design-Testing" will permit the design, testing and validation of models. It will consist of the area where the models can be designed, and will contain the design tools, the representation tools, the control tools (necessary to run the model), and the meta-analysis tools. In order to design a model, students will have to determine the model' s entities, their properties and the relations between them. A library will provide two kind of entities: concrete ones (that correspond to objects); and abstract ones (that may correspond directly to concepts). For each entity, one or more properties will have to be determined as well as the estimation of their values (if necessary). The entities will be connected through relations in order to build the model. The software will provide different kind of relations: Qualitative logical relations; qualitative semantic relations; semi-quantitative relations; quantitative simple algebraic relations. Once the model has been created, the student will then be able to run it. Before or after model running, the student can activate the Representation Tools, including: tables of data, graphs, bar charts, and decision table associated with the desirable entities' properties. It has to be pointed out that the table of data is an "open" one; that means it permits the entrance of external numerical data. This characteristic enables the user to represent data from external real experiments, and compare them with the data coming from the model, -a necessary step in the phase of model's validation. Meta-analysis tools: These tools permit to students to see in a visual or verbal mode the historic of their modelling process.
3. The "Structured Notebook" will stimulate students to take notes during the modelling process.
4. The "Help system" will contain three sub-components: an on-line help (a hypertext system), an immediate context sensitive help giving audio guidelines on using the various components, and an intelligent help system, which will be activated during specific interface actions (during "decision making" modelling).
5. The "Files' management system": will implement a special view of the file system, hiding the file's real location on the hard disk and showing the hierarchy of the classes and their students/users instead. In addition, a special tree is 'assigned' to every user (student or teacher), which will represent the various themes of study and the underlying created models. As a result, all a user needs to know to access his created models is their name in the system, the class they belong to, and the specific subject of study they are interested in.
6. The "Teacher Assistant Tools": Three different tools will permit teachers correspondingly to add or delete users, to observe in details the student's activities historic and to allow him/her to create and insert new entities.

It has to be noted that a prototype of this system has already developed (in C++,Win32 API) and experimented in real school life situations, funding by a national initiative program. Consequently, the component "Study Themes Creation Tool", "Structured Notebook" & "Files' management system" are already developed, while the main tools of the "Models Design-Testing" need to be seriously extended. Other tools such as the Meta-analysis tools and the Teacher Assistant's Tools have to be created from scratch

ii) Innovative aspects concerning the Learning effective distance Communication/Collaboration

Our approach is based on two main assumptions:

1. We consider that a distance-learning site must engage students into concrete learning activities rather than simply provide information as many information-based sites do, considering teaching to be a process of transmitting information.
2. In order to have real effective communication, encouraging cooperation and collaboration between students as well as between students and teachers, appropriate learning activities and tools permitting the support collaborative reasoning both for students and teachers, are needed.

In order to allow and support teaching and collaborative learning through the Web-Based MODELLINGSPACE, we need two categories of tools and services addressed to students and teachers, distinguished to conventional & non-conventional ones:

A. Conventional tools and services a.Provided to students: basic educational material about modelling and models in sciences; links to other interesting sites; sites presenting the schools and students workgroups involved; discussion forums; whiteboard for announces; model publishing facilities b.Provided to teachers: educational material (list of situations to study, pedagogical and teaching alternative approaches, students' difficulties); discussion forums; FAQs

B. Non conventional tools and services: Assuming the importance of well specifying collaborative settings, it is useful to mention some well-considered examples of them, before presenting the corresponding tools. Collaborative learning schemes among students:

1. "Negotiation of differences": The case when students negotiate their differences in the modelling approach on the same "problem". The students having worked on the problem in their private workspace, then having made their solutions public, the system recognises the 'interesting' significant differences of the various solutions provokes the students, directly or via the teacher, to confront their solutions.

2. "Collaborative common modelling process": The case when the students work on partial models of a complex "problem" and then collaborate in order to produce the global model of the "problem" (e.g. a concept map of a complex environmental situation). The same collaborative scheme may include the case when the students start on-line to create a model of a "problem" from scratch.

3. "Student-student tutoring": The case when a student is engaged in a modelling process and another student assists him, making suggestions and giving advices. The pair of students is arranged or suggested by a teacher, between students of the same age as well as of different ages, a learning situation that can be mutually fruitful for the participants.

4. "Learning from examples": The case when a student wants to study already published and commented models of "problem" produced by other students. The student can study an example by one or more of the following modes: a) The student sees the model of the "problem", produce explanatory notes and then compares his notes with these of the model's creator, b) The student can see the model creation's historic in visual mode and/or the notes accompanying it. In both modes, the student can communicate with the model's creator without a teacher's 'presence' being required.

5. "Apprenticeship style of learning ": The case when a well-experienced student is engaged in a modelling process in public, and another student observes this process. The advanced student must be able to answer the observer's questions, as well as to explain and justify his answers.

It must be noticed that the evolution of a collaboration process, starting from a given scheme, may trigger the activation of another scheme.

a. Various tools will be provided to students (Student Assistance Tools) to support the collaboration between students:

• Presence info tool: provides information about the users being actually on-line. A 2D spatial metaphor can be used for presenting students and teachers from each school and also a videoconference application will be available.
• Public workspace tool: permits to declare one's individual workspace as a public one.
• Meeting planner tool: permits to be registered in a collaborative learning or teaching scheme and indicates the beginning of the activity
• Awareness of others activity tool: provides information on others' activity
• Communication tools: permit to communicate with the off-line (asynchronous communication) and on-line (synchronous communication) members of the learning community for multiple reasons, before entering into a concrete learning activity.
• Collaboration-dialogue based tools: For each of the different collaborative learning schemes and the teaching schemes a 'specific collaboration supported dialogue tool' can be activated, linked with a 'chat':
• Co-ordination tool of the shared workspace (modelling design): permits to manage the users and the actions taken place in the shared workspace during model's creation.
• Students' activity meta-analysis tool: permits to present an elaborated historic of the learning activity (collaborative or teaching), in the appropriate visual mode, promoting the development of meta-conceptual awareness of students.

b. Various tools will be provided to teachers (Teacher Assistance Tools):

• Presence info tool
• Meeting planner tool
• Communication tools: (similar to these of students).
• Awareness of students' activity tools: permits to teacher to receive information about the user_id, the task he works on, the collaborative scheme selected, to watch every student's public-individual workspace and to get elaborated comparative information concerning the various workspaces (e.g. students' models, presenting interesting differences).
• Students' work analysis tool: presents an elaborated analysis (historic) of the students' actions in the shared workspace, during model's creation or model's modification, carrying out also a comparison of students models.
• Visualisation of global student activity tool: permits to present the participants' time-sequence process (both dialogues and shared workspace actions) in an appropriate visual mode.
• General activity Record Database: permits global recording of the whole on-line actions/interactions/collaborations of teachers and students, permitting queries. A set of agents and sub-agents is necessary in order to have available and functional the above described tools.

ii) Innovative aspects concerning the educational research

The MODELLINGSPACE development requires successive evaluation with a small or more important number of students, while its application in real school settings will have to cope with: a) four educational systems and curricula, b) a wide range of students' ages, c) different subject matters.

In all the cases, the methodological approach will mainly based on ethnographic (systemic) and discourses analytic methods focused on significant qualitative data, during formative successive evaluations. This 'heavy' large-scale qualitative methodological approach will be carried out for first time in a European level, in order to support the development and the implementation of an innovative technology-based learning environment.

The research approach will be based on the collection and analysis of Various Kinds of Data:

• The data collection will be based on participant and non participant observation of students actions and talk, in school
• Detailed observation data will be collected including video and audio and verbatim transcriptions, ¨
• Messages and dialogues via Internet, will be stored and analysed
• Significant actions and interactions between students-students and students and teachers on ModellingSpace interface, will also be stored and analysed
• Student learning in different "social" systems will be observed, i.e. a) in the classroom, in single groups working with the research-teacher and individually, b) in different 'social systems' formed working by distance through Internet
• Pre-post test tasks will also be given for a focused identification of concept formation.
• In parallel, the teachers of the school community, having working with the ModellingSpace will be also interviewed.

Teachers participating in the research will have two different profiles: a) these participating from the beginning to the ModellingSpace project, and b) these incorporating ModellingSpace activities, influenced only by the software itself, the accompanied material and some seminar on ModellingSpace project.

The research will use in a relatively extensive way discourse analytic methods for two kinds of data:

a) Data on natural oral dialogues taking place in classroom

b) Data on written dialogues and exchanges through tools permitting or supporting dialogue via Internet

The discourse analysis will be based on identification, and categorisation of different "speech acts", according to different intentions of interlocutors. The analysis will use specific approaches according to different nature of dialogue (see for instance: Sabah, Dimitracopoulou and all 2000, for natural oral dialogue, and Quignard & Baker 1999, for Internet based written kind of dialogues).

It has to be noted that before the implementation of innovation to the experimental schools, extensive and successive evaluations on Human Computer Interaction through the interface of ModellingSpace environment will be done, with the participation of individual students and group of students in laboratory settings (into the context of specific Human Computer Interaction Laboratories).

For the various possible interactions and tasks permitted by ModellingSpace, an appropriate combination of methods and tools will be used (such as: the Usability Analyzer (UA), the Hierarchical Task Analysis, and the Goals-Operators-Methods-Selection Rules (GOMS).

Users of ModellingSpace

ModellingSpace is addressed to 11 to 17-year-old students.

During the system development, experimental research will be carried out. The first users will be therefore the students of the schools where these experiments will be carried out, namely:

- In Greece: ,
- Experimental Primary School Primary School Secondary (12-15)
- 3rd Gymnasium of Rhodes (Veneto-kleion Gymnasium) Gymnasium Secondary (15-18)
- 1st Lykeio of Rhodes, (Venetokleion Lykeio) Lycée

- In Belgium:
- Centre éducatif communal secondaire - La Garenne - Charleroi - General and technical secondary (12-19)
- Institut Sainte-Marie - La Louvière - General Secondary (12-18)
- Institut technique de la Communauté française - Val d'Escaut - Antoing - General and technical (12-16)

- Institut provincial des Arts et Métiers (IPAM) - Nivelles - General and technical secondary (12-19) 

- In France:
- Ecole de Seiches sur le Loir Primary school
- Ecole Chateauneuf sur Loir Secondary (12-15) College
- Classes d'applica-tion de l'Institut Universitaire de Formation des Maitres, des pays de Loire Secondary (15-18) Lycee

- In Portugal:
- Escola Secundaria do Monte de Caparica Secondary (12-18)
- Escola Basica 2,3 Sophia de Mello Breyner - Brandoa Basic (10 - 15)

The experimental research during system development will be carried out in primary and high schools in national levels and will be focused on multinational experimental application on distance collaborative learning and teaching in high schools.

Deliverables

Work plan

The project work plan calls for an in-depth review of the state of the art about the concept of modelling and of modelling tools, design of the pedagogical approach and of the research frame, establishment of a human and technical network, design and achievement of teachers training, computational modelling tool development and Internet integration, collection and analysis of empirical data, development of pedagogical methodologies and teachers guide lines and dissemination and exploitation actions.

For these purposes, the work has been organised in the following work-packages:

WP 1 Project management, coordination, evaluation and quality assurance
OBJECTIVES:
- To specify the project management structures concerning the communication and quality assurance.
- To determine and maintain the financial issues.

WP 2 Pedagogical Approach and Research Methodology
OBJECTIVES:
- To achieve in depth review on model's concept, modelling process and types of modelling.
- To carry out critical analysis of existing computerized modelling tools.
- To fulfil pedagogical approach conception and research methodology design.
- To attain modelling processes and development of high-level cognitive skills.
- To determine user requirements in computational modelling processes.

WP 3 Pedagogical Scenarios and Activities - Set up the School and Human Network
OBJECTIVES:
- To analyse and compare the national curricula of Belgium, French, Portugal, Greece
- To design the pedagogical scenarios and activities.
- To determine the user requirements in the framework of MODELLINGSPACE.
- To set up human and technical network.
- To determinate communication procedures.
- To set up and establish communication channels with teachers participating in the project and school networks.
- To provide to teachers education on the concept of modelling and other project's specific elements.
- To design teachers' training model
- To develop the teachers' training material and the manuals.

WP 4 Design and Development of MODELLINGSPACE
OBJECTIVES:
- To analyse and select the distance learning and agent platforms to be used.
- To specify, design and implement the MODELLINGSPACE software.
- To facilitate formative and summative evaluation of the MODELLINGSPACE software and to set up testing of MODELLINGSPACE under real learning situations.

WP 5 Empirical Research: Collection and analysis of data - Evaluation of MODELLINGSPACE
OBJECTIVES:
- To accomplish research in the framework of schools participating in the project.
- To realise data analysis.
- To proceed to software evaluation.

WP 6 Information dissemination and Exploitation plan
OBJECTIVES:
- To promote the project concepts and solutions for widespread adoption.
- To elaborate an exploitation plan for the deployment of ModellingSpace application and commersialisation.