lecture15

# R g0 is unknown data n independent draws of

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

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Example II: Coverage probability of mean in a response model when the standard deviation is unknown, and the distribution is Gaussian Model: Y = µ + R, .....R ~ G(0,σ ) , σ is UNKNOWN Data: n independent draws of r.vs Y1, Y2, € …Yn Observations: y1,y2,…yn Result: ~ T= µ− µ ~ σ n −1 ~ t n −1 n € Proof: See notes from class Algorithm: Step 1: From the t ­table, Find c such that P( ­c< tn ­1<c) =0.95 Step 2: The coverage interval is ( ~ ~ σ n −1 ~ σ n −1 µ− c , µ+ c ) n n ~ € Step 3: The confidence interval is ? Other results derived from the proof (i) The reason for sample s.d being divided by n ­1 (ii) Confidence interval for σ2. Definition: Pivotal Quantity Case III: Confidence interval for a Binomial model Case IV: Confidence interval for a regression model...
View Full Document

## This note was uploaded on 09/27/2012 for the course STAT 231 taught by Professor Cantremember during the Winter '08 term at Waterloo.

Ask a homework question - tutors are online