STAT 344 Chapter Notes - Chapter unit2: Sampling Distribution, Scatter Plot, Datapoint
6 Feb 2018
School
Department
Course
Professor
Document Summary
Unit #2: ratio and regression estimation, inference for domains. We started this unit with the notion that sometimes a helper variable could be our friend. If (i), we get to record both y and x for sampled units (x is the helper variable), and (ii), xp, the population mean of x, is a known quantity, then there is scope to proceed. Both the obvious (or vanilla") estimator, ys, and the ratio estimator, ( ys/ xs) xp, are defensible estimators of the target, yp. Through a combination of intuition, simulation work, and mathematics, we established the following. Aside: let"s always be mindful of what we mean when we say that estimator 1 performs better than estimator 2. This is a remark about the sampling distributions of the two estimators - namely that the rst has a sampling distribution that is more concentrated around the target than the second.
