Moscow Modelling Colloqium

THE FUTURE OF MODELLING

Richard Millwood

New Learning Environments

Anglia Higher Education College

December 5th 1990

INTRODUCTION

This paper is intended as an introduction to a session about the future of modelling. Modelling in this paper means the activity of producing a computer executable representation of some system and the consequent evaluation of that representation by, for example, using the model purposefully and comparison with a real (or imagined) system. The paper addresses the following questions in order to provoke discussion:

1 Why do we want to make modelling an activity for learners?

2 What are the computer based modelling techniques for classrooms in the future?

3 What modelling skills do students need?

4 How do students acquire modelling skills?

5 How should we assess modelling skills?

6 What further research is needed?

1 WHY DO WE WANT TO MAKE MODELLING AN ACTIVITY FOR LEARNERS?

A discussion of the future of modelling cannot take place without some preliminary thoughts on why we want students to do modelling. These thoughts may then guide us as new opportunities arise and research questions are posed. My thoughts (in order of importance) are as follows: Students should do modelling in order:

- to understand the basis for decision making in a democratic society;

- to increase their knowledge and understanding in a range of subject matters;

- to improve their ability to solve problems, take decisions and think about the problem solving process;

- to enable direct involvement as managers or politicians in the process of decision making and predicting outcomes;

- to understand, communicate and solve scientific and social problems as researchers.

Clearly not all students can become expert modellers, but this does not mean that they can know nothing of models or the uses to which they are put. This kind of understanding is perhaps most important for the majority and should occupy much more of our attention.

2 WHAT ARE THE COMPUTER BASED MODELLING TECHNIQUES FOR CLASSROOMS IN THE FUTURE?

Current modelling techniques are largely numerical, quantitative, based on mathematical calculus and use time as an independent variable, in other words dynamic modelling.

The future of computer based modelling techniques for classrooms might also include a range of different kinds:

- logical techniques such as those used in expert systems;

- techniques for ordering factors to support decision making;

- techniques of describing relationships between factors in a qualitative manner;

- techniques for relating variables based on position as an independent variable in one, two or three dimensions.

The above list is made from the point of view of kinds of models. It is also important to consider future modelling techniques in terms of a learner's progression and learner's knowledge and experience. There are techniques in modelling which do not relate directly to the kind of modelling, but are general to all kinds of modelling. For example:

"identification of the important factors"

or

"deciding what is cause and what is effect."

3 WHAT MODELLING SKILLS DO STUDENTS NEED?

In order to answer this question we could make lists of modelling skills which relate to types of model. For example, in dynamic modelling, the skill of making appropriate mathematical expressions. But there are also meta-level skills, such as maintaining a sense of an overall modelling process and how to carry on when "stuck". I believe that having skills in the latter category is essential for success, whereas more support can be gained from the computer for the skills of expressing models.

4 HOW DO STUDENTS ACQUIRE MODELLING SKILLS?

There are few explicit teaching programmes for "how to model". Often students are simply shown to finished models, taught some "toolkit" skills and left to get on with it.

In future we may improve on this state of affairs by:

- encouraging reflection on the modelling process;

- discussing case studies;

- extending teachers' understanding of the conceptual needs.

As before I believe it is essential that educators have a clearer picture of modelling at all stages and levels of thought employed in the construction and evaluation of a model.

5 HOW SHOULD WE ASSESS MODELLING ACTIVITIES?

Modelling is frequently an open ended activity. It is not clear when the student has come to a sensible finishing point, since by definition refinements are always possible. Nevertheless, we can look at three aspects of modelling activities in order to make assessments. On the one hand we can examine the outcome of such work by asking questions such as:

can a decision be made?;

can future patterns be predicted?.

Alternatively we can examine the process and development of a model and consider the student's methods and strategy. Finally we can ask the student to make self-evaluative reports on their work.

6 WHAT FURTHER RESEARCH IS NECESSARY?

It seems to me that we need to know more clearly what the range of techniques of modelling are, how they are employed and at what developmental stage can students be introduced to concepts in modelling. Research should be perhaps be directed at the following:

- learner's preconceptions of models if they have them;

- the ways in which expert modellers approach their task;

- the conceptual framework required for modelling at all levels;

- the teaching strategies currently employed where modelling is taught as a subject.