Solutions_to_Assignment_5

Solutions_to_Assignment_5 - Stat 332/322 Short Solutions to...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
Stat 332/322 Short Solutions to Assignment 5 1. (a) ˆ μ ( y ) ratio = ˆ θμ ( x ) = 2 . 70, where ˆ θ = ˆ μ ( y ) / ˆ μ ( x ). The estimated variance of the cor- responding estimator is d V ar μ ( y ) ratio ) = (1 - n/N σ 2 /n = 0 . 0034 where ˆ σ 2 = i s ( y i - ˆ θx i ) 2 / ( n - 1) = 0 . 037. 1. (b) fit a linear model Y i = α + β ( x i - ¯ x ) + R i , then we have ˆ α = 3 . 2 and ˆ β = 1 . 02. Then the regression estimate is ˆ μ ( y ) reg = ˆ α + ˆ β * ( μ ( x ) - ¯ x ) = 3 . 2 + 1 . 02 * (2 . 43 - 2 . 88) = 2 . 74 and the estimated variance is d V ar μ ( y ) reg ) = (1 - n/N ) * ˆ σ 2 /n = 0 . 0021, where ˆ σ 2 = i s ( y i - ¯ y - ˆ β * ( x i - ¯ x )) 2 / ( n - 1) = 0 . 023. 1. (c) For the particular data we have here, the sample size is a bit of too small, and both ratio and regression estimators can be unreliable. If we pass this issue, then the general guideline in choosing between ratio and regression estimator is to look at the scatter plot: if a fitted regression line has a close-to-zero intercept, then ratio estimator is preferred, due to its simplicity; otherwise the regression estimator should be used. The regression estimator
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 01/12/2012 for the course STAT 332 taught by Professor Xu(sunny)wang during the Fall '09 term at Waterloo.

Page1 / 3

Solutions_to_Assignment_5 - Stat 332/322 Short Solutions to...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online