Abstract
The methods of trend surface analysis, a form of regression analysis, are applied to complex archaeological surfaces, or response surfaces, derived from a series of points representing carved stone monuments in the Lowland Classic Maya region bearing the most recent dates. Results of a comparison of trend surface and residual maps produced and of an analysis of variance with expected maps do not support a strong west to east diffusionary trend for the collapse, nor a hypothesized invasion. Evidence is presented indicating the probable existence of five regional zones or sociopolitical spheres in the Late Classic whose existence may bear directly on the collapse through increased competition. The utility of the trend surface model is demonstrated, primarily as a method of formalized data description and as an aid in the building of process-response models.

This publication has 5 references indexed in Scilit: