Reducing Large Data Sets Using Granger Causality: A Paleoecological Example from the Columbia Plateau

Summary

This is an abstract from the "SAA 2019: General Sessions" session, at the 84th annual meeting of the Society for American Archaeology.

This paper presents proxies of vegetation, climate, human population, and fire from late Holocene sediments from the Columbia Plateau (USA). Statisitical analyses such as multivariate regression and Granger Causality time series analysis are used to reduce complexity and illuminate the underlying structure of the data set. Results show that multivariate regression can overfit a model in such cases, identifying variables as statistically significant when they are not. Granger causality removes multicollinearity and identifies a smaller set of more tightly interconnected variables, revealing basic lead-lag relationships over several time intervals. Differences in conclusions based on the two statistical procedures are discussed, along with the implications of choosing one method over the other.

This Resource is Part of the Following Collections

• Society for American Archaeology 84th Annual Meeting, Albuquerque, NM (2019)

Keywords

Record Identifiers

Abstract Id(s): 24308