The case of partial least squares (PLS) path modeling in managerial accounting research

Abstract

Managerial accounting researchers are often challenged to create sophisticated path models to answer research questions. Because of their specific characteristics, partial least squares (PLS) path modeling offers a wide range of useful possibilities for accounting scholars. Nevertheless, PLS path models remain an underutilized analytical tool in managerial accounting research. One reason for their underutilization may be that there has been no systematic discussion of PLS path modeling in accounting that draws on the newest findings. Therefore, we discuss the characteristics of PLS path models, such as the use of composite factors for construct measurement, the explorative characteristics of PLS for path modeling, the relevance of prediction orientation for practical research, and introduce tools such as mediation analysis, heterogeneous data modeling, and importance-performance matrix analysis. Overall, this paper facilitates the adoption of PLS path models by updating the current conventional understanding of PLS path models in managerial accounting.