ABSTRACT
This paper is concerned with developing approaches to limits of confidence in output attribute values given errors in the input attribute values. To do this we construct methods to measure the variation in the attribute values obtained from suitability analyses given variations in the attributes of the input source maps that comprise the suitability analysis of interest. Typically, it is very difficult to measure the actual errors associated with input maps. Therefore, to get an idea of how the resultant map attributes change with respect to variations of the attributes in the input map, perturbations of the attributes associated with the input maps can be imposed. We do not study variation due to perturbation of shapes of the polygons, only confidence limits in geographic suitability analysis due to input attribute error, both real and imposed. Errors in the shapes of the polygons themselves have been studied by Maffini et al. (1989), ªObservations and comments on the generation and treatment of error in digital GIS dataº (this volume) and by Honeyc utt (1986).
