UCSB Anthropology Brown Bag Lecture Series Presents:

The Use of GIS to Test for Bias in Non-random surveys: A Southwest German Example

By Susan Harris

Although the use of statistical methods has dictated that a random sample is the ideal, in many parts of the world a random survey design is not possible. In these cases, archaeological survey should not be abandoned. Instead, methods should be devised to explore how far the actual survey differs from a hypothetical random sample. A GIS database offers the level of data storage and retrieval to carryout such a comparison using standard test statistics such as teh Student's-t and Chi-square, as well as a simulation approach, to pinpoint sources of bias for various landform variables. A case-study from Stone Age Southwest Germany illustrates these principles.



Wednesday April 16, 1997
HSSB 2001A, The Anthropology Conference Room

	Wish to contact this person? Send a note through e-mail.
	Back to list of Brown Bag talks
	Back to main UCSB Anthropology page