 




nearest neighbour analysis 





A method for comparing the distribution of points in an area with a theoretical norm based on their random distribution within the same area. The theoretical distribution gives an expected average distance between all pairs of nearest neighbours, which is compared with the observed distribution for the mapped pattern. The test statistic Rn is derived by dividing the observed by the expected average distance, and ranges from 0.0 (indicating that the observed distribution is more clustered than expected from a random allocation of the points in that space), through 1.0 (a random distribution), to 2.149 (a more uniform distribution than expected): the scale of values is not linear, however (i.e. a value of 0.4 doesn\'t imply a distribution twice as clustered as that shown by a value of 0.8), which creates interpretative difficulties. The size of Rn for any set of points is also influenced by the size of the area analysed: what is clustered at one scale may be uniform at another.Â (RJJ)
Suggested Reading Aplin, G.J. 1983: Orderneighbour analysis. Concepts and Techniques in Modern Geography 36. Norwich: Geo Books. 





Bookmark this page:






<< former term 

next term >> 









Other Terms : rural geography  Fordism  structuralism 
Imprint 
Home 
Add new article 
Your List 
Tools 
Become an Editor 
Tell a Friend 
Links 
Awards 
Testimonials 
Press 
News 
About

Copyright ©2009 GeoDZ. All rights reserved. Terms of Use  Privacy Policy  Contact Us
