ECON 3210 Lecture 13: Econ 3210 Lecture 13
7 Jan 2018
School
Department
Course
Professor
Document Summary
The sample average has two very nice properties: 1. These two concepts are both desirable properties of an estimator, but they are distinct (i. e. , an estimator can be biased but consistent, biased and inconsistent, unbiased and consistent, or biased and inconsistent) Consistency: in large enough samples, we"ll get it right . Unbiasedness: if we keep drawing new samples and constructing yn, we would be right, on average. clear all. *execute the program 1000 times, saving mean and variance each time set seed 1234 simulate y_bar=r(mean) sigma2=r(var), reps(1000) nodots: econ3210_p1. Consistency follows by the lln, which we discussed last class. The law of large numbers (lln) says that, as a sample gets arbitrarily large, the sample mean will become arbitrarily close to the average in the population. An estimator is consistent when it converges to the true population parameter (that is is designed to estimate) as the sample size grows arbitrarily large.
