R: genModelString

genModelString

R Documentation

genModelString

Description

Creates lavaan model strings from model matrices.

Usage

genModelString(
  Lambda = NULL,
  Phi = NULL,
  Beta = NULL,
  Psi = NULL,
  Theta = NULL,
  tau = NULL,
  Alpha = NULL,
  useReferenceIndicator = !is.null(Beta),
  metricInvariance = NULL,
  nGroups = 1
)

Arguments

`Lambda`	Factor loading matrix.
`Phi`	Factor correlation (or covariance) matrix. If `NULL`, all factors are orthogonal.
`Beta`	Regression slopes between latent variables (all-y notation).
`Psi`	Variance-covariance matrix of latent residuals when `Beta` is specified. If `NULL`, a diagonal matrix is assumed.
`Theta`	Variance-covariance matrix of manifest residuals. If `NULL` and `Lambda` is not a square matrix, `Theta` is diagonal so that the manifest variances are 1. If `NULL` and `Lambda` is square, `Theta` is 0.
`tau`	Intercepts. If `NULL` and `Alpha` is set, these are assumed to be zero.
`Alpha`	Factor means. If `NULL` and `tau` is set, these are assumed to be zero.
`useReferenceIndicator`	Whether to identify factors in accompanying model strings by a reference indicator (`TRUE`) or by setting their variance to 1 (`FALSE`). When `Beta` is defined, a reference indicator is used by default, otherwise the variance approach.
`metricInvariance`	A list containing the factor indices for which the accompanying model strings should apply metric invariance labels, e.g. `list(c(1, 2), c(3, 4))` to assume invariance for f1 and f2 as well as f3 and f4.
`nGroups`	(defaults to 1) If > 1 and `metricInvariance = TRUE`, group specific labels will be used in the measurement model.

Value

A list containing the following lavaan model strings:

`modelPop`	population model
`modelTrue`	"true" analysis model freely estimating all non-zero parameters.
`modelTrueCFA`	similar to `modelTrue`, but purely CFA based and thus omitting any regression relationships.