Learn R Programming

robcat (version 0.2)

polyserial_efficiency: Efficiency of minimum DPD estimators of polyserial model

Description

Calculate population asymptotic variance-covariance matrix associated with a parameter vector theta, assuming that the polyserial model is correctly specified and that theta is the true model parameter. May take a few moments to compute because a relatively large number of integrals need to be numerically solved.

Usage

polyserial_efficiency(theta, alpha)

Value

A numeric matrix that is the population asymptotic variance-covariance matrix associated with a parameter vector theta and tuning constant alpha.

Arguments

theta

Parameter vector of polyserial model; assumed to be the true one. First element is polyserial correlatio coefficient, second is the population mean of X, third is population variance of X, and the remaining elements are the thresholds associated with the ordnal Y (must be in increasing order)

alpha

Tuning constant governing robustness-efficiency tradeoff. Set to 0 for maximum likelihood.

Examples

Run this code
theta <- c(rho = 0, mu = 0, sigma2 = 1, tau1 = 0) # true parameter vector
polyserial_efficiency(theta, alpha = 0.5)

Run the code above in your browser using DataLab