Estimation issues with PLS and CBSEM: Where the bias lies! ☆

By Marko Sarstedt, Joseph F. Hair, Christian M. Ringle, Kai Oliver Thiele, Siegfried P. Gudergan. 2016

1,709 citations recorded by the source corpus.

Discussions concerning different structural equation modeling methods draw on an increasing array of concepts and related terminology. As a consequence, misconceptions about the meaning of terms such as reflective measurement and common factor models as well as formative measurement and composite models have emerged. By distinguishing conceptual variables and their measurement model operationalization from the estimation perspective, we disentangle the confusion between the terminologies and develop a unifying framework. Results from a simulation study substantiate our conceptual considerations, highlighting the biases that occur when using (1) composite-based partial least squares path modeling to estimate common factor models, and (2) common factor-based covariance-based structural equation modeling to estimate composite models. The results show that the use of PLS is preferable, particularly when it is unknown whether the data's nature is common factor- or composite-based.

Subfields: Artificial Intelligence, Computer Networks and Communications

View Paper: https://www.semanticscholar.org/paper/66571b7d97764d32aebcff6abbd54ed36186e8ed

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Networks and Communications

Citations in the corpus

Assessing measurement model quality in PLS-SEM using confirmatory composite analysis (2020, methodology): “fluenced, affected, or caused by the underlying latent variable (Sarstedt et al., 2016).”
ASSESSING REFLECTIVE MODELS IN MARKETING RESEARCH: A COMPARISON BETWEEN PLS AND PLSc ESTIMATES (2018, methodology): “…data from common factor populations, PLSc performs well in the situation of having more indicators (≥ 4 indicators) in a reflective measurement model and also having larger sample sizes (n ≥ 250) in a study, to avoid peak result in terms of coefficients' mean absolute error (Sarstedt et al., 2016).”
The influence of a digital strategy on the digitalization of new ventures: The mediating effect of digital capabilities and a digital culture (2021, methodology): “See Hair et al. (2016) and Sarstedt et al. (2016) for a detailed description of the PLS method and its usage.”
Mirror, mirror on the wall: a comparative evaluation of composite-based structural equation modeling methods (2017, methodology): “^ To address this call and further the emancipation of composite-based SEM from its factor-based sibling, researchers require a more nuanced understanding of these methods’ performances based on strong simulation results (Henseler et al. 2014; Sarstedt et al. 2016).”
Partial least squares structural equation modeling in HRM research (2020, methodology): “PLS-SEM enables researchers to address a broad range of research questions by (1) allowing for estimating formatively specified measurement models (Sarstedt et al., 2016), (2) handling non-normal data (Cassel, Hackl, & Westlund, 1999), and (3) working effectively with a wide range of sample sizes (H”