Researching Geography: The Indian context by Gopal Krishan & Nina Singh

Author:Gopal Krishan & Nina Singh
Language: eng
Format: mobi
ISBN: 9781138206397
Publisher: Taylor and Francis
Published: 2016-11-02T18:30:00+00:00

Clustered

Contagious spatial diffusion

Houses in an agglomerated settlement

Random

Chance spatial occurrences

Early settlements in a virgin region

Uniform

Competitive spatial situations

Central places as in Christaller’s model

All distributional patterns belong either to one of the three types mentioned above or to millions of patterns lying between clustered and random or random and uniform. Nearest neighbour statistic provides a precise measure of any distribution pattern. It also indicates the possible underlying factors that were instrumental in generating a particular pattern.

In the case of competitive spatial situations, a uniform pattern of distribution will be created. Such is a case often with distribution of agricultural market towns in an area. This technique is particularly useful for identifying distortions in the distribution of public facility points, such as schools, hospitals and post offices. It is observed that in India, the facilities provided by the central government, such as post offices, are more evenly distributed, than that provided by the state governments due to greater political interference in the latter case. Guidelines for future planning can be provided by indicating the locations that could fill the gaps.

The nearest neighbour statistic summarises the spatial picture of the whole map. This can be done using geographical analysis of machine (GAM) developed by Openshaw et al. (1987) and refined by Fotheringham and Zhan (1996). Otherwise, the conventional method is to compute the mean of the distances, point by point, between all the objects in spatial distribution and fix it in the formula:

R n = 2 d N / A , where Rn is the nearest neighbour statistic, d is the mean distance between the nearest neighbours, N is the number of objects in the distribution and A is the area of the region.

The computed statistic will range from zero, if the distribution is clustered, to 2.15 if the distribution is uniform or completely ordered. If a value of 1.00 is obtained then the pattern is called ‘random’. If the Rn statistic is 1.8, it means that the distribution has a tendency towards uniformity, and if is 0.8 it has a tendency towards randomness (generally speaking).

The nearest neighbour statistic is easy to calculate and comprehend. The shape of the region affects the statistic and its associated test of significance. For example, if the region’s shape is long, narrow and/or rectangular, it may have relatively low values of R. Here points are by necessity close to one another. Another situation that makes a difference to the analysis is the location of points close to the boundary of the study region. Consider the case of port towns in a coastal region. The book Quantitative Methods in Geography by P.J. Taylor (1977) is the best for learning the technique of calculating Rn statistics.

Using the quadrat method, nearest neighbour test and the Moran statistic, the researcher obtains a single measure of net pattern for a map consisting of point locations. This single summary score is often called the ‘global statistic’. The null hypothesis is: there is no underlying pattern, or deviation from randomness, among the set of points.

In other

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.