Covariance Matrix of the EBA Utility Scale

Computes the (normalized) covariance matrix of the utility scale from the covariance matrix of elimination-by-aspects (EBA) model parameters.

cov.u(object, norm = "sum")

Arguments

object	an object of class eba, typically the result of a call to eba
norm	either sum (default), a number from 1 to number of stimuli, or NULL; see uscale for details

Details

The additivity rule for covariances \(cov(x + y, z) = cov(x, z) + cov(y, z)\) is used for the computations.

If norm is not NULL, the unnormalized covariance matrix is transformed using \(a^2 cov(u)\), where the constant \(a\) results from the type of normalization applied.

Value

The (normalized) covariance matrix of the utility scale.

See also

uscale, eba, wald.test.