Conventional test of measurement invariance in three steps

Computes a set of CFA models (e.g., configural, metric, and scalar/threshold invariance) and compares them to each other.

Usage

globalMI(
  ...,
  chi.sq = FALSE,
  omit = "",
  what = c("cfi", "tli", "rmsea", "srmr", "chisq"),
  partial = NULL,
  more = NULL,
  ho = F,
  quiet = FALSE
)

Arguments

...: Formula, group, and all the other eligible arguments of cfa function.
chi.sq: Logical, if chi-square should be included in the model comparison table.
omit: Character (vector). Which model should be omitted. Possible values "configural", "metric", "scalar".
what: Character (vector) of fit measures to report (see fitMeasures
partial: List of parameters to release constraints, see `group.partial` lavOptions
more: More models with extra constraints, currently limited to `residuals` and `means` which are added on the top of scalar model (or the most constrained otherwise, if 'scalar' is omitted)
ho: Currently defunct, for the higher-order factor models invariance see globalMIH
quiet: Logical, if the results should be printed.

Value

A list with all the fitted models, an LRT test lavTestLRT, and a table of fit indices (`what`) and their differences across fitted models.

Details

If the model has ordered indicators, then instead of configural-metric-scalar it fits configural-threshold-scalar sequence of models.