Within multivariate regression problems influence analysis and leverage analysis have been performed, namely based on a generalized least squares (GLS) estimation of regression coefficients. In cases where the covariance matrix is unknown multistage estimation procedures have been studied.
Measures of influence and leverage for a three-step estimator that uses the residuals from an initial ordinary least squares (OLS) regression to estimate the covariance matrix of the response vectors are given, namely (a) the derivative influence and (b) the deletion influence. Both of these measures can be represented as the sum of components that correspond to the stages of the GLS estimation. Usefulness of the chosen compromise influence measure has been demonstrated by an example.