matrixpls (version 1.0.13)

cei: Composite Equivalence Indices

Description

The matrixpls method for the generic function cei computes composite equivalence indices (CEI) for the matrixpls object. By default, the composites are compared against unit-weighted composites.

Usage

cei(object, object2 = NULL, ...)

Arguments

object

matrixpls estimation result object produced by the matrixpls function.

object2

Another matrixpls object that matrixpls is compared against.

...

All other arguments are ignored.

Value

Composite equivalence indices as a vector

Details

Composite equivalence indices quantify if two sets of composites calculated from the same data using different weight algorithms differ. Composites are matched by name and correlations for each pair are reported.

See Also

Other post-estimation functions: ave(), cr(), effects.matrixpls(), fitSummary(), fitted.matrixpls(), gof(), htmt(), loadings(), predict.matrixpls(), r2(), residuals.matrixpls()

Examples

# NOT RUN {
# Load the Tenenhaus et al 2005 model and data from semPLS
library(semPLS)
data(ECSImobi)
data(mobi)

# Reflective and empty formative model
reflective<- ECSImobi$M
formative <- t(reflective)
formative[] <- 0

# Estimation using covariance matrix
model <- list(inner =  t(ECSImobi$D),
              reflective = reflective,
              formative = formative)


S <- cor(mobi)

matrixpls.ModeA <- matrixpls(S, model, innerEstim = innerEstim.centroid)
matrixpls.Fixed <- matrixpls(S, model, weightFun = weightFun.fixed)

cei(matrixpls.ModeA)
cei(matrixpls.ModeA, matrixpls.Fixed)
# }