"Traditional" SEM versus piecewise SEM -Costs and Benefits
I recently came across an article of Lefcheck (2016) covering an alternative way of evaluating SEM models, namely piecewise SEM
Link: https://arxiv.org/ftp/arxiv/papers/1509/1509.01845.pdf
The article makes some interesting points on the potential benefits of piecing multiple models together versus traditional SEM that calculates parameter estimates based on the whole observed variance-covariance matrix. For instance, it releases restrictions of the "t-rule" (atleast 5 observations per parameter estimate at the highest-level) and can handle non-normal distributed data.
After some searching I found that piecewise sem is most often used in ecological / evolution / biology type of research. However, I cant find any study in organizational behavior literature or organizational psychology literature (my field of research)
Because I am not an expert in SEM I would love to hear your opinion on this approach, especially regarding 1) is anyone familiar with this approach and have used it in a paper and 2) could it potentially also be usefull for mulitlevel organizational research or is there a drawback to this type of analyses for these kind of settings?
Interested in hearing your thoughts!
Other references:
Shipley, B. (2000). A New Inferential Test for Path Models Based on Directed Acyclic Graphs. Structural
Equation Modeling, 7, 206–218.
Shipley, B. (2009). Confirmatory path analysis in a generalized multilevel context. Ecology, 90, 363–368.
Shipley, B. (2013). The AIC model selection method applied to path analytic models compared using a dseparation
tests. Ecology, 94, 560–564.