Chapter 15 Approximating the distributions of econometric estimators and test statistics

摘要: Publisher Summary Approximate distribution theory derives results from assumptions on the stochastic process generating data. The quality of approximation is not better than specifications which it based. models used by econometricians are, at best, crude and rather arbitrary. As most methods employ information first four moments data whereas usual asymptotic typically requires only two moments, some loss in robustness must be expected. However, if a rough idea about degree skewness kurtosis available, that can often exploited to obtain considerably improved approximations sample statistics. chapter discusses sophisticated appropriate situations where econometrician able make correct detailed being studied. In current practice, applied occasionally draw incorrect conclusions basis alleged properties their procedures. recent years, an extraordinary fondness for has developed among econometricians. Considerable effort devoted showing new estimator or test asymptotically normal efficient. assertion given approximately suggests speaker believes would sensible treat as though were really normal. Accurate convenient distributions econometric estimators statistics are great value.