Abstract:
Sample selection models or selectivity models are often used to analyse structural
and evaluation studies. Everywhere, individuals who are selected into one group,
will likely be different from the characteristics of individuals in other groups. This
selection problem is the core problem that must be considered. As this selection
process occurs, this is a very important factor to be observed. Ignoring the non-
random selection will lead to selection bias, either on the basis of observed or
unobservable characteristics which will be the focus of this study.
Sample selection models have been widely used in various fields since the
introduction of a two step estimator called Heckman’s Two Step Estimator.
However, this model is only concern on the participant, either in parametric, semi-
parametric and non-parametric sample selection models. Hence, in this study,
parametric and semi-parametric sample selection models are proposed. Instead of the
participant as in previous models, the proposed models focus on non-participant for
both parametric and semi-parametric of sample selection models.
