"failure to identify the function of a model construct and failure to proceed accordingly has been at the root of many methodological controversies in geography"

Harvey (1969) Explanation in Geography

Research Overview
During the 1950s and 1960s geography went through what has become known as its quantitative revolution. During this time the scientific method was emphasized, with the intention of pushing geographical investigation beyond description of static patterns and towards establishing relationships between measures of those patterns and their change. As geography moved towards become a law-making discipline, mathematical and statistical models were adopted. During 1970s however, many human geographers became dissatisfied with the approach these models and methods implied.

During my study, development and application of several types of quantitative and simulation model - including statistical, cellular automata, individual-based and agent-based - I’ve become interested in some of the philosophical issues surrounding these tools. For example, agent-based simulation tools developed since the end of geography’s quantitative revolution seem to potentially overcome some of the criticisms of statistical and mathematical approaches. The epistemological roles of these simulation tools are potentially both heuristic and dialogic.

Papers on this aspect of my work include Millington et al. (2011) looking at participatory evaluation of agent-based land use models, Millington and Perry (2011) reviewing multi-model inference methods in biogeography, Perry and Millington (2008) examining spatial modelling of succession-disturbance dynamics in terrestrial ecological systems and Millington et al. (2006) considering methods for quantifying wildfire regimes.

As well as checking out those papers, you can use read entries on my blog discussing modelling and philosophy to get a sense of where I'm coming from (but note when the entry was written - I may have moved on from what I was thinking at that time...).