Skip to main content Skip to navigation

Computers for Learning: An Empirical Modelling perspective

Abstract

In this thesis, we explore the extent to which computers can provide support for domain learning. Computer support for domain learning is prominent in two main areas: in education, through model building and the use of educational software; and in the workplace, where models such as spreadsheets and prototypes are constructed. We shall argue that computer-based learning has only realised a fraction of its full potential due to the limited scope for combining domain learning with conventional computer programming. In this thesis, we identify some of the limitations in the current support that computers offer for learning, and propose Empirical Modelling (EM) as a way of overcoming them.

We shall argue that, if computers are to be successfully used for learning, they must support the widest possible range of learning activities. We introduce an Experiential Framework for Learning (EFL) within which to characterise learning activities that range from the private to the public, from the empirical to the theoretical, and from the concrete to the abstract. The term 'experiential' reflects a view of knowledge as rooted in personal experience. We discuss the merits of computer-based modelling methods with reference to a broad constructionist perspective on learning that encompasses bricolage and situated learning. We conclude that traditional programming practice is not well-suited to supporting bricolage and situated learning since the principles of program development inhibit the essential cognitive model building activity that informs domain learning. In contrast, the EM approach to model construction directly targets the semantic relation between the computer model and its domain referent and exploits principles that are closely related to the modeller's emerging understanding or construal. In this way, EM serves as a uniform modelling approach to support and integrate learning activities across the entire spectrum of the EFL. This quality makes EM a particularly suitable approach for computer-based model construction to support domain learning.

In the concluding chapters of the thesis, we demonstrate the qualities of EM for educational technology with reference to practical case studies. These include: a range of EM models that have advantages over conventional educational software due to their particularly open-ended and adaptable nature and that serve to illustrate a variety of ways in which learning activities across the EFL can be supported and scaffolded.