This is an R-Package for generating instance-specific cut-off values for structural equation models (SEM). Cutoffs are generated by simulating data from either an original dataset or a population model provided in lavaan syntax.
For a more detailed description of the packages functionalities see the integrated help files at CRAN and the GitHub repository.
There you can also find the latest development version.
The catalyst for this novel approach was the realization that predetermined model fit cutoff values such the ones suggested by Hu and Bentler (1999) and Schermelleh-Engel, Moosbrugger, and Müller (2003) are not sufficiently generalizable across model constellations and sample characteristics.
Previous research has demonstrated a wealth of biasing factors: sample size, measurement quality, model complexity, magnitude and location of misspecification
(Heene, Hilbert, Draxler, Ziegler, & Bühner, 2011; Iacobucci, 2010; Marsh, Hau, & Wen, 2004; McNeish, An, & Hancock, 2018; Moshagen & Auerswald, 2018; Nye & Drasgow, 2011)
The ezCutoffs package generates cutoff values for the exact model and sample constellation under consideration. It does so by generating a large number of random data sets using the parameters of the empirical data as a template.
Once fitted, the 95th (or 90th, or 99th, ...) percentiles of the resulting distribution of fit indices can be used as empirically justified cutoff values to which the empirical fit can be compared.