The Nearest-Neighbor Statistic: Archaeological Application and New Developments

Abstract

Human behavior has spatial corollaries of great importance to both geographers and archaeologists. Some of these corollaries can be expressed as point maps, and to describe them objectively it has become common-place to employ the nearest-neighbor analysis technique devised by ecologists. But recently it has been shown that "underestimation" is an inherent flaw of the technique. Initially, therefore, the attention of this paper focuses on elimination of this problem by modifying the traditional formulae. With the flaw corrected, the issues surrounding interpretation of the nearest-neighbor statistic are discussed, and a review is made of the special difficulties of applying the technique to archaeological data.