simsem (version 0.2-8)

simMisspecSEM: Set of model misspecification for SEM model.

Description

This function will define model misspecifiction from a defined model. This function is similar to simSetSEM such that the matrices that indicates misspecification will be added as arguments in the function. However, users do not have to add all matrices and vectors in the function. Only element indicating misspecification is added.

Usage

simMisspecSEM(..., exo = FALSE, conBeforeMis=TRUE, misBeforeFill=TRUE, 
misfitType="rmsea", misfitBound=new("NullVector"), averageNumMisspec=FALSE, 
optMisfit="none", numIter=20)

Arguments

...

Arguments definition is listed in the Details section of simSetSEM. Again, this function does not require to list all required matrices or vectors like the

exo

specify TRUE if users wish to specify both exogenous and endogenous indicators.

conBeforeMis

TRUE if users wish to constrain parameters before adding misspecification. FALSE if users wish to constrain parameters after adding misspecification.

misBeforeFill

TRUE if users wish to apply misspecification before applying the auto-completion on the parameters that users have not specified. FALSE if users wish to apply the auto-completion before adding misspecification. This option is hel

misfitType

The type of population misfit used in the misfitBound below. The default is "rmsea". The two other options are "f0" and "srmr". See popMisfitMACS fo

misfitBound

The lower and upper bounds of the population misfit. This option must be a vector with two elements.

averageNumMisspec

If TRUE, the misfit will be divided by the number of free elements in the misspecification object. The default is FALSE.

optMisfit

Use the optimization method to pick the misspecification set. That is, the program will draw a number of misspecification sets. Then, the different sets of misspecification will be compared together. If "min" is specified, the program will pi

numIter

The number of different misspecification sets for comparison in the optimization method.

Value

• object in SimMisspec that saves model misspecification.

See Also

• simSetSEMfor matrix definition and how to specify SEM model
• SimMisspecfor the simResult
• simMisspecCFAfor misspecification model in CFA andsimMisspecPathfor misspecification model in Path analysis.

Examples

u2 <- simUnif(-0.2, 0.2)
n1 <- simNorm(0, 0.1)
loading.X.trivial <- matrix(NA, 6, 2)
loading.X.trivial[is.na(loading.X.trivial)] <- 0
LX.trivial <- simMatrix(loading.X.trivial, "u2")
error.cor.X.trivial <- matrix(NA, 6, 6)
diag(error.cor.X.trivial) <- 0
RTD.trivial <- symMatrix(error.cor.X.trivial, "n1")
error.cor.Y.trivial <- matrix(NA, 2, 2)
diag(error.cor.Y.trivial) <- 0
RTE.trivial <- symMatrix(error.cor.Y.trivial, "n1")
RTH.trivial <- simMatrix(matrix(NA, 6, 2), "n1")
SEM.Mis.Model <- simMisspecSEM(LX = LX.trivial, RTE = RTE.trivial, RTD = RTD.trivial, RTH = RTH.trivial, exo=TRUE)