Munich Personal RePEc Archive

Synthetic data: an endogeneity simulation

Carbajal De Nova, Carolina (2014): Synthetic data: an endogeneity simulation.


Download (11MB) | Preview


This paper uses synthetic data and different scenarios to test treatments for endogeneity problems under different parameter settings. The model uses initial conditions and provides the solution for a hypothetical equation system with an embedded endogeneity problem. The behavioral and statistical assumptions are underlined as they are used through this research. A methodology is proposed for constructing and computing simulation scenarios. The econometric modeling of the scenarios is developed accordingly with the feedback obtained from previous scenarios. The inputs for these scenarios are synthetic data, which are constructed using random number machines and/or Monte Carlo simulations. The outputs of the scenarios are the model estimators. The research results demonstrated that a treatment for endogeneity can be developed as the sample size increases.

MPRA is a RePEc service hosted by
the Munich University Library in Germany.