An Introduction to Spatial Pattern Analysis in a GIS Environment

An Introduction to Spatial Pattern Analysis in a GIS Environment
University of California, Santa Barbara CA
14-18 August 2000

Instructors:
Arthur Getis (Workshop Coordinator)
Michael Goodchild
Lauren Scott
John Weeks

Host institution:
CSISS, University of California, Santa Barbara

This workshop focuses on applications of pattern analysis in a Geographic Information Systems environment. It features 1) a series of illustrated lectures on both GIS and spatial pattern analysis, 2) exercises demonstrating the principles outlined in the lectures, and 3) data exploration based on current projects concerned with spatial patterns relating to various social, behavioral, and economic phenomena. The workshop lectures will emphasize the fundamental principles and examples of the use of spatial pattern analysis for the help it gives toward the solution of important societal questions. Demonstrating concepts covered in the lectures, exercises will utilize a variety of software tools including ArcView, S-Plus, Point Pattern Analysis, and clustering software developed by Lauren Scott (Scott Software). The data exploration portion of the workshop will consist of GIS-based analyses of spatial data related to 1) the incidences of various crimes within an urban environment, 2) the spread of infectious diseases in the tropics, 3) the diffusion of fertility decline in a third world setting, and 4) variables selected by the participants prior to the workshop. Participants will prepare a final report of their GIS-based analyses.

Workshop Leaders:
Arthur Getis, Michael Goodchild, Lauren Scott, and John Weeks will conduct the workshop. Professor Getis, San Diego State University, has had a long association with spatial statistics, especially in the area of point pattern analysis. Perhaps his best known work is the collaborative effort in which he has been engaged with Professor J. Keith Ord (currently of Georgetown University) on the development of a series of mainly locally based spatial association statistics. Michael Goodchild is best known for his work in geographic information systems and digital libraries. He is Professor of Geography at UC Santa Barbara, Director of CSISS, and Associate Director of the Alexandria Digital Library. Dr. Scott recently received her Ph.D. in geography in the joint doctoral program of UC Santa Barbara and San Diego State University. Environmental Sciences Research Institute currently employs her in Redlands, California. She is expert in the development and use of various aspects of GIS. Her dissertation employed the use of spatial association statistics in a GIS environment to gauge the effect of changing workplace and residential location on urban spatial structure using as her case study detailed data from the Los Angeles Metropolitan Area. Professor Weeks, San Diego State University, is a well-known demographer and author of the standard textbook in the field, now in its 7^th edition. His recent work includes the study of spatial patterns of fertility variables in an Egyptian context (to be published in Environment and Planning A) and spatial patterns of crime in a GIS environment. Professor Weeks has broad experience in the social sciences, having taught in both the Departments of Geography and Sociology at SDSU.

Pedagogical Goals:

The workshop will be geared toward PhD candidates and young faculty members in the social sciences and health related disciplines. The purposes are not only to inform those new to spatial pattern analysis about its uses, but also to instill in them a fundamental understanding of the importance of spatial thinking and spatial concern. In the spirit of CSISS, we hope to “facilitate intensive cross-fertilization of research ideas…” among a diverse group from fields such as political science, criminal justice, epidemiology, sociology, and anthropology. The workshop will attempt to extend the analytical power of spatial analysis to social science. Participants will be encouraged to bring with them a georeferenced data set in which they have a particular interest. They will be informed beforehand of possible web sites where such data are available. Data will be made available to those without personal data sets. Participants will take part in the development and discussion of a report that they create on the use of spatial pattern analysis.

Calendar:

Monday, 14 August:

Lectures:

Exercise:       Operating K Function Analysis (Scott Software)
Orientation:    Model Development in Pattern Analysis
Application:   Developing Hypotheses

CORE PROGRAMS > National Workshops > Spatial Pattern Analysis

Last Update: February 7, 2001