The use of partial least squares path modeling in causal inference for archival financial accounting research

Abstract

In financial accounting research, multivariate regression is almost exclusively the dominant statistical method. By contrast, Partial Least Squares path modeling is a under-utilized statistical method. The aim of this study is to examine how Partial Least Squares path modeling can be applied to the archival financial accounting research. This article first presents an overview on multivariate regression and structural equation modeling. The authors then highlight that advantages of using Partial Least Squares path modeling to address the research constraints in causal inference for archival financial accounting research.