A more general linear aggregation model is considered in that we allow for collîne- arity between the independent or explanatory variables. Thus the analysis is presented in a framework utilizing Moore-Penrose generalized inverses of singular matrices.