Composite reliability (Raykov's rho / McDonald's omega)

Computes composite reliability (Raykov's rho / McDonald's omega) for a set of congeneric reflective indicators: $$\rho = \frac{(\sum \lambda_i)^2 \text{Var}(\eta)}{(\sum \lambda_i)^2 \text{Var}(\eta) + \sum \theta_i}$$ where \(\lambda_i\) are the unstandardized factor loadings and \(\theta_i\) are the indicator error (unique) variances.

Usage

drm_raykov_rho(M, loadings = NULL, error_variances = NULL, factor_var = 1)

Arguments

M

A numeric matrix or data frame of indicators (rows = observations, cols = indicators).

loadings

Optional named numeric vector of indicator loadings.

error_variances

Optional named numeric vector of unique error variances.

factor_var

Variance of the latent construct (default 1.0).

Value

Numeric composite reliability scalar in [0, 1] (or NA_real_ if not computable).

References

Raykov T (1997). “Estimation of Composite Reliability for Congeneric Measures.” Applied Psychological Measurement, 21(2), 173–184. doi:10.1177/01466216970212006 .

McDonald RP (1999). Test Theory: A Unified Treatment. Lawrence Erlbaum Associates, Mahwah, NJ.