tut2_10

tut2_10 - - prediction interval for a new response at x is...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
The University of Western Ontario Department of Statistical and Actuarial Sciences Statistical Sciences 3859a Tutorial 2 1. Consider the regression through the origin model: y i = βx i + ε i for i = 1 , 2 , . . . , n and where the ε i ’s are independent normal random variables with mean 0 and variance σ 2 . (a) Derive the Least-Squares estimator for β : b β . (b) Obtain the unbiased estimator for σ 2 : MSE. (c) Show that b β is unbiased. (d) Show that the variance of b β is σ 2 n i =1 x 2 i and give an estimator for the standard error of b β . (e) Show that if x 0 is a given number, then b y | x 0 = b βx 0 is an unbiased estimator for E [ y | x 0 ] = βx 0 . (f) Find the variance of b y | x 0 , and deduce that a 1 - α confidence interval for E [ y | x 0 ] is b βx 0 ± t α/ 2 ,n - 1 v u u t MSE( x 0 ) 2 n i =1 x 2 i . (g) Note that a 1
Background image of page 1
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: - prediction interval for a new response at x is b x t / 2 ,n-1 v u u t MSE 1 + x 2 n i =1 x 2 i ! . 2. Do not use lm() in this question. Consider the following data: x y 1 2 2 5 3 6 4 8 5 9 (a) Fit the regression through the origin model and estimate the error variance. (b) Compute the standard error estimate of b and obtain a 95% condence interval for . (c) Compute a 95% condence interval for E [ Y | x = 4 . 5]. (d) Give a 95% prediction interval for a new response at x = 4 . 5. 3. Repeat the previous question using the R function lm() ....
View Full Document

This note was uploaded on 11/25/2010 for the course AS 3859 taught by Professor Braun during the Fall '10 term at UWO.

Ask a homework question - tutors are online