PSTAT 109 Lecture 6: Lecture 6 Statistical Inference
3 Sep 2016
School
Department
Course
Professor
Document Summary
6 population , sample x, s independent and identically distributed, random sample consistent: estimator is approximately = to the parameter it estimates for a large sample size. X ~ n( x, ) mean, variance as n is large, 2 decreases confidence interval methods invnorm method: known. B = invnorm (1-tail, x, sqrt(): calculator: known (given stats, calculator: unknown (given stats) Use s instead of so t-distribution! , x, n, confidence level (sometimes s as well) given: z interval. S, x, n, confidence level given: t interval. Clt n large x normal regardless of original distribution regardless of original distribution. Unknown (t interval) selection of sample size (n) when estimating . Null hypothesis (h0): (not) enough evidence to reject null hypothesis. Is innocent / fair contains the information which gives the benefit of the doubt as to whether is true or false. H0: = 0 vs. h1: < 0. H0: = 0 vs. h1: > 0.
