getInvarCompTable: Compare model fit of nested dyadic invariance models in order...

In jsakaluk/dySEM: Dyadic Structural Equation Modeling

Compare model fit of nested dyadic invariance models in order from most parsimonious (residual) to least parsimonious (configural)

Description

Compare model fit of nested dyadic invariance models in order from most parsimonious (residual) to least parsimonious (configural)

Usage

getInvarCompTable(mods)

Arguments

mods	a list of neted lavaan dyadic invariance models, in the order of residual, intercept, loading, configural

Value

a data frame of model fit statistics for each model, as well as the difference in fit statistics between each model and the previous model

Examples

dvn <- scrapeVarCross(dat = commitmentQ, x_order = "spi",
x_stem = "sat.g", x_delim1 = ".", x_delim2="_", distinguish_1="1", distinguish_2="2")

sat.residual.script <- scriptCFA(dvn, lvname = "Sat",
constr_dy_meas = c("loadings", "intercepts", "residuals"), constr_dy_struct = "none")

sat.intercept.script <- scriptCFA(dvn, lvname = "Sat",
constr_dy_meas = c("loadings", "intercepts"), constr_dy_struct = "none")

sat.loading.script <- scriptCFA(dvn, lvname = "Sat",
constr_dy_meas = c("loadings"), constr_dy_struct = "none")

sat.config.script <- scriptCFA(dvn, lvname = "Sat",
constr_dy_meas = "none", constr_dy_struct = "none")

sat.residual.fit <- lavaan::cfa(sat.residual.script, data = commitmentQ,
std.lv = FALSE, auto.fix.first= FALSE, meanstructure = TRUE)

sat.intercept.fit <- lavaan::cfa(sat.intercept.script, data = commitmentQ,
std.lv = FALSE, auto.fix.first= FALSE, meanstructure = TRUE)

sat.loading.fit <- lavaan::cfa(sat.loading.script, data = commitmentQ,
std.lv = FALSE, auto.fix.first= FALSE, meanstructure = TRUE)

sat.config.fit <- lavaan::cfa(sat.config.script, data = commitmentQ,
std.lv = FALSE, auto.fix.first= FALSE, meanstructure = TRUE)

mods <- list(sat.residual.fit, sat.intercept.fit, sat.loading.fit, sat.config.fit)

getInvarCompTable(mods)

jsakaluk/dySEM documentation built on March 18, 2024, 1:01 p.m.