Testing with Exponentially Tilted Empirical Likelihood

Imposing restrictions without assuming underlying distributions to modelize complex realities is a valuable methodological tool. However, if a subset of restrictions were not correctly specified, the usual test-statistics for correctly specified models tend to reject erronously a simple null hypothesis. In this setting, we may say that the model suffers from misspecification. We study the behavior of empirical phi-divergence test-statistics, introduced in Balakrishnan et al. Statistics 49:951–977 (2015), by using the exponential tilted empirical likelihood estimators of Schennach Ann Stat 35:634–672 (2007), as a good compromise between the efficiency of the significance level for small sample sizes and the robustness under misspecification.