Rural Settlement in England: Analysing Environmental Factors and Regional Variation in Historic Rural Settlement Organisation Using Regression and Clustering Techniques

Author(s): Andrew Lowerre

This report shows how it is possible to collate Geographic Information Systems (GIS) data for historic settlement nucleation and dispersion with a range of data on environmental variables in order to investigate the relationships between them. Ordinary Least Squares (OLS) regression model specification, selection and validation procedures, followed by further analysis using spatial regression methods, identified environmental variables that appear to have had the most significant influence on settlement organisation. The use of OLS and spatial regression and the innovative Relative Area Overlap (RAO) technique has enabled investigation of how relationships between key environmental variables and historic settlement organisation varied across England. Overall, the regression analyses indicate that far more of the variation in the measures of settlement organisation is not explained by the environmental variables than is explained by them. The results of the RAO analysis echo this conclusion. Using unsupervised classification, it has been possible to develop new, national-scale characterisations of historic settlement organisation and of key environmental variables. These new classifications of historic settlement organisation often broadly align with Brian Roberts and Stuart Wrathmell’s delineations of provinces, sub-provinces and local regions, but the cluster outlines and Roberts and Wrathmell’s boundaries diverge more often than they agree.

Report Number:
72/2014
Series:
Research Report
Pages:
152
Keywords:
Desk Top Assessment Digital Data Environmental Studies Medieval Modern Post Medieval Area Assessment Geographical Information System

Accessibility

If you require an alternative, accessible version of this document (for instance in audio, Braille or large print) please contact us:

Customer Service Department

Telephone: 0370 333 0607
Email: [email protected]

Research