El experimento natural como un nuevo diseño cuasi-experimental en investigacion social y de salud

Abstract

Currently in social science research it is used a new research design defined as a "natural experiment", where the researcher has none participation in the treatment assignment because it has been directly done for an exogenous agent such as a law, a social policy or a natural event. This implicit exogenous intervention allows that low-level pre-determined selection bias groups be formed. Therefore, the comparison on results being found between these groups (law-affected group, social policy-affected group, or natural event-affected group versus contrafactic groups not being affected) allow to predict the effect of this law, social policy, or natural event almost so efficiently like happen in randomly created treatment-and-control groups.