Educational technology has yet to deliver the beneﬁts or successes that were expected in educational practice, especially in relation to issues other than the communication and delivery of teaching materials. Evidence suggests that these diﬃculties stem from the mismatch between formalised virtual learning environments and everyday sense-making and between the rich potential for enhanced learning aﬀorded by new technology and the constraints of old-style educational practice. In addressing this mismatch, some commentators suggest that the primary need is for a new culture of educational practice—and even that such a culture is already emerging, and others identify the need for a new paradigm for educational technology. The aim of this thesis is to explore the potential for a new paradigm for educational technology based on the principles and tools of Empirical Modelling (see http://dcs.warwick.ac.uk/modelling).
The thesis builds upon previous research on Empirical Modelling as a constructionist approach to learning, and in particular Roe’s doctoral thesis ‘Computers for learning: an Empirical Modelling perspective’. Roe’s treatment of Empirical Modelling can be viewed as generalising the use of spreadsheets for learning through applying ‘programming by dependency’ within the framework of existing educational practice. In contrast, this thesis is concerned at a more fundamental level with the contribution that Empirical Modelling can make to technology enhanced learning that may lead to new educational practices. In particular, it identiﬁes eight signiﬁcant characteristics of learning that are well-matched to Empirical Modelling activity, and associates these with experimental, ﬂexible and meaningful strands in learning. The credentials of Empirical Modelling as a potential new foundation for educational technology are enhanced by demonstrating that Empirical Modelling is radically diﬀerent from traditional software development and use. It provides a methodology for modelling with dependency that is more closely related to the use of spreadsheets for learning.
The thesis elaborates on the relationship between Empirical Modelling and learning in a variety of diﬀerent contexts, ways and applications. Three examples drawn from computer science higher education are explored to emphasise the experimental, ﬂexible and meaningful characteristics of Empirical Modelling. This discussion of Empirical Modelling in a speciﬁc educational context is complemented by an investigation of its relevance to learning in a wider context, with reference to a broad range of sub jects, to speciﬁc issues in language learning, and to the topics of lifelong learning and collaborative learning. Although the application of Empirical Modelling for learning is as yet too immature for large scale empirical studies, its potential is evaluated using informal empirical evidence arising from Empirical Modelling practice at Warwick. The sources for this evaluation are well-established teaching activities relating to Empirical Modelling in Computer Science at the University of Warwick, comprising an introductory module and a number of ﬁnal year undergraduate pro jects.
The thesis concludes by considering the extent to which Empirical Modelling can go beyond the support for constructionism envisaged by Roe, to address the broader agenda of supporting constructivist learning using computers. To this end, a close relationship between Empirical Modelling and a vision of constructivism recently set out by Bruno Latour in his paper ‘The Promises of Constructivism’ is demonstrated.