Problems in Location Sampling

ABSTRACT In geographical and related research uncertainty and error have arisen in the selection, application, and interpretation of designs in plane sampling largely through a failure to differentiate between area sampling and location sampling procedures. A survey of some commonly used area sampling designs indicates the frequency and extent of this error. Location sampling has remained a poorly understood technique through its confusion with area sampling, and there has been a consequent failure to grapple with the problems of stratifying by location when the primary sampling items are of varying areal extent or are irregularly spaced. An objective method of stratifying by location employing minimal aggregation is described and is used as a basis for both systematic and random samples. Tests on locationally stratified farm sample designs, based upon certain known and hypothetical farm characteristics in one New South Wales shire, indicate that increases in sampling precision obtained through locational stratification are directly related to the levels of spatial segregation in the items being sampled.