plspath: Estimation of pls-path models

In cbsem: Simulation, Estimation and Segmentation of Composite Based Structural Equation Models

Description

plspath estimates pls path models using the classical approach formulated in Lohmueller.

Usage

plspath(dat, B, indicatorx, indicatory, modex = "A", modey = "A",
  maxiter = 100, stdev = FALSE)

Arguments

dat	(n,p)-matrix, the values of the manifest variables. The columns must be arranged in that way that the components of refl are (absolutely) increasing
B	(q,q) lower triangular matrix describing the interrelations of the latent variables: b_ij= 1 regression coefficient of eta_j in the regression relation in which eta_i is b_ij= 0 if eta_i does not depend on eta_j in a direct way (b_ii = 0 !)
indicatorx	(p1,1) vector indicating with which exogenous composite the x-indicators are related.
indicatory	(p2,1) vector indicating with which endogenous composite the y-indicators are related. The components of the indicators must be increasing.
modex	equals "A" or "B" , the mode for this block of indicators
modey	equals "A" or "B" , the mode for this block of indicators
maxiter	Scalar, maximal number of iterations
stdev	Boolean Should the standard deviations of the estimates be computed by bootstrap?

Value

out list wih components

Bhat	(q,q) lower triangular matrix with the estimated coefficients of the structural model
eta	(n,q)-matrix, the scores of the latent variables
w	vector of length p of weights for constructing the latent variables
lambdahat	vector of length p with the loadings
resa	(n,?) matrix of residuals from outer model
resi	(n,?) matrix of residuals from inner model
R2	vector with the coefficients of determination for all regression equations of the structural model
iter	number of iterations used
ret	scalar, return code:
	0 normal convergence
	1 limit of iterations attained, probably without convergence
sdev.beta	(q,q) matrix, the standard deviations of path coefficients (when stdev = TRUE)
sdev.lambda	vector, the standard deviations of loadings (when stdev = TRUE)

Examples

data(mobi250)
refl <- c(1, 1, 1, 4, 4, 4, 2, 2, 2, 3, 3, 5, 5, 5, 6, 6, 6, 7, 1, 1, 4, 4, 4, 4)  
o <- order(refl)
dat <- mobi250[,o]
dat <- dat[,-ncol(dat)]
refl <- refl[o][-length(refl)] 
indicatorx <- refl[1:5]
indicatory <- refl[-c(1:5)] - 1
B <- matrix(c(0,0,0,0,0,0,1,0,0,0,0,0,0,1,0,0,0,0,
              0,1,1,0,0,0,0,1,1,1,0,0,1,0,0,0,1,0),6,6,byrow=TRUE)  
out <- plspath(dat,B,indicatorx,indicatory,modex="A",modey="A")

cbsem documentation built on May 2, 2019, 5:56 a.m.