STAT 4201 Midterm: STAT 421 OSU Point Estimation II Complete

• School: Ohio State University
• Department: Statistics
• Course: STAT 4201
• Professor: All

Document Summary

Note: highlighted parts show the differences from 4/1 version, and underlined phrases were emphasized in the class. Idea of consistency: as sample size n of a rs from some population with distribution f x increases to infinity, does the estimator converge to , i. e. , that the error goes to zero in some sense. Probability, i. e. , the absolute error converges to zero in probability. ] In case of standard distributions, usually this is not a problem. But there are always pathological cases, in which chance variation may produce surprise outcomes, so we must be watchful about the effect of these pathologies. But as n increases, outliers can take us away from the true value, with a positive probability, exposing us to a very large risk, and possibly lead us to bankruptcy! Cauchy( ,1), where is small (of the order of 10. [note: is the median of the distribution , not. 0 , e(x) does not exist, and var(x) is not finite.