genPCA {cds}

Generate PCA data and Calculates Correlation Matrices

Description

Generate a response style data set from a specific correlation matrix, clean the data with constrained dual scaling and report the original, cleaned and contaminated correlation matrices in a list.

Usage

genPCA(nr.indv = rep(100, 5), m = 10, q = 7, r = 3, err.coeff = 0.1,
  alphamat = rbind(c(0.5, 2, 4), c(10, 2, 10), c(1, 2, 1), c(4, 2, 0.5),
  c(0.1, 2, 0.1))[1:length(nr.indv), ], randomize = TRUE, ...)

Arguments

nr.indv

Vector; number of individuals in each response style group. It is passed to simpca.

m

scalar; Number of items.

q

scalar; Number of rating categories, such that the rating scale is 1:q.

r

scalar; Rank of simulated correlation matrices.

err.coeff

scalar; Standard deviation used in simulations that is passed on to simpca.

alphamat

matrix; Contains the spline parameters for the different response styles that is passed to simpca.

randomize

logical; See simpca.

...

Arguments passed to cds.

Value

A list with components:

Rsim

Correlation matrix from which the sample was generated

Rclean

Correlation matrix for the cleaned data

Rcont

Correlation matrix for the contaminated data

Author(s)

Pieter C. Schoonees