Composite-based partial least squares structural equation modeling (PLS-SEM) has become a well-established element in researchers’ multivariate analysis methods toolbox (Hair, Black, Babin, & Anderson 2018). Particularly PLS-SEM’s ability to handle highly complex path models and its causal-predictive nature, which allows bridging the apparent dichotomy between explanation and prediction, have contributed to its massive dissemination. While its usage spans across multiple fields outside the social sciences, the mainstay of PLS-SEM is business research. Some of the most popular models in the fields – including customer satisfaction and loyalty models (e.g., Ahrholdt, Gudergan, & Ringle 2019), corporate reputation models (e.g., Hult, Hair, Proksch, Sarstedt, Pinkwart, & Ringle 2018), and technology acceptance models (e.g., Schubring, Lorscheid, Meyer, & Ringle 2016) – are routinely estimated using PLS-SEM. It is not surprising that some of the most cited articles in the Journal of Business Research (JBR) use the PLS-SEM method (e.g., Coltman, Devinney, Midgley, & Venaik 2008; Camisón & Villar-López 2014).
Recent research has brought forward numerous methodological extensions that allow for a more nuanced assessment of results. These extensions include, for example, latent class segmentation, model comparisons, endogeneity assessment, and predictive model evaluation (Hair, Hult, Ringle, & Sarstedt 2017; Hair, Sarstedt, Ringle, & Gudergan 2018). Especially the prediction-oriented PLS-SEM analyses (Shmueli, Ray, Velasquez Estrada, & Chatla 2016; Sharma, Shmueli, Sarstedt, Danks, & Ray 2019) and methods to assess the result’s robustness (Sarstedt, Ringle, Cheah, Ting, Moisescu, & Radomir 2019) are particularly important to substantiate findings, conclusions, and managerial recommendations.
The aim of this special issue of JBR is to introduce advanced PLS-SEM methods to a wider audience. The special issue embraces the applications of advanced PLS-SEM methods to generate new insights and shed new light on existing models and theories. In addition, methodological advances of the PLS-SEM method will also be considered. Potential topics include, but are not limited to:
- Differences in model development from explanatory vs. predictive perspectives,
- Explanatory versus predictive model evaluation,
- New metrics for goodness-of-fit testing and predictive power assessment,
- Using PLS-SEM in experimental research and on experimental data (e.g., discrete choice modelling data),
- Endogeneity in PLS-SEM,
- Common method variance in PLS-SEM,
- Using PLS-SEM with archival (secondary) data,
- Addressing observed (multi-groups analysis and moderation) and unobserved heterogeneity (segmentation) in PLS-SEM,
- Using PLS-SEM on panel or longitudinal data,
- Combining Bayesian modeling and PLS-SEM, and
- Other advanced developments of PLS-SEM and their application
Resource: Journal Elsevier.