I found a bug in the nullRMSEA function. For this is that the moreFitIndices was giving me problems
The function calls lavaan::lav_partable_independence(fit), where it should be lavaan::lav_partable_independence(object). So, the moreFitIndices function works with the example code because the example code creates an lavaan object called "fit", keeps using the same object called fit instead of the given lavaan object
Sorry, I didnt make it a pull request, still havent got to learn to use github to those features
nullRMSEA
function (object, scaled = FALSE, silent = FALSE)
{
fit <- update(object, model = lavaan::lav_partable_independence(fit)) ### chage fit for object
fits <- lavaan::fitMeasures(fit, fit.measures = c("rmsea",
"rmsea.scaled", "rmsea.robust"))
if (scaled) {
RMSEA <- fits["rmsea.robust"]
if (is.na(RMSEA))
RMSEA <- fits["rmsea.scaled"]
if (is.na(RMSEA))
RMSEA <- fits["rmsea"]
}
else RMSEA <- fits["rmsea"]
if (!silent) {
cat("The baseline model's RMSEA =", RMSEA, "\n\n")
if (RMSEA < 0.158) {
cat("CFI, TLI, and other incremental fit indices may not be very",
"informative because the baseline model's RMSEA < 0.158",
"(Kenny, Kaniskan, & McCoach, 2015). \n")
}
}
invisible(RMSEA)
}
Hi
I found a bug in the nullRMSEA function. For this is that the moreFitIndices was giving me problems
The function calls lavaan::lav_partable_independence(fit), where it should be lavaan::lav_partable_independence(object). So, the moreFitIndices function works with the example code because the example code creates an lavaan object called "fit", keeps using the same object called fit instead of the given lavaan object
Sorry, I didnt make it a pull request, still havent got to learn to use github to those features
nullRMSEA function (object, scaled = FALSE, silent = FALSE) { fit <- update(object, model = lavaan::lav_partable_independence(fit)) ### chage fit for object fits <- lavaan::fitMeasures(fit, fit.measures = c("rmsea", "rmsea.scaled", "rmsea.robust")) if (scaled) { RMSEA <- fits["rmsea.robust"] if (is.na(RMSEA)) RMSEA <- fits["rmsea.scaled"] if (is.na(RMSEA)) RMSEA <- fits["rmsea"] } else RMSEA <- fits["rmsea"] if (!silent) { cat("The baseline model's RMSEA =", RMSEA, "\n\n") if (RMSEA < 0.158) { cat("CFI, TLI, and other incremental fit indices may not be very", "informative because the baseline model's RMSEA < 0.158", "(Kenny, Kaniskan, & McCoach, 2015). \n") } } invisible(RMSEA) }