Fryges, Helmut; Wagner, Joachim - Forschungsinstitut zur Zukunft der Arbeit <Bonn> - 2007
pre-treatment variables X, if
(1) ()d D X for all dY ⊥∈D ,
with as the outcome associated with treatment level d. It is …-treatment variables X may be difficult when the
dimension of X is large. In the case of binary treatment, Rosenbaum und Rubin (1983 … f dx= .
Then the generalised propensity score is defined as R = r(D, X) (Hirano and Imbens 2004: 2).
Assuming that the …