Dr. Hackney STA Solutions pg 194

# Dr. Hackney STA Solutions pg 194 - 12-4Solutions Manual for...

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: 12-4Solutions Manual for Statistical InferenceFor the logit model we haveEstimateStd. Errorz valuePr(>|z|)Intercept15.04297.37192.041.0413temp-.2322.1081-2.147.0318and for the probit model we haveEstimateStd. Errorz valuePr(>|z|)Intercept8.770843.862222.271.0232temp-.13504.05632-2.398.0165Although the coefficients are different, the fit is qualitatively the same, and the probabilityof failure at 31◦, using the probit model, is.9999.12.19 a. Using the notation of Example 12.3.1, the likelihood (joint density) isJYj=1eα+βxj1 +eα+βxjy*j11 +eα+βxjnj-y*j=JYj=111 +eα+βxjnjeα∑jy*j+β∑jxjy*j.By the Factorization Theorem,∑jy*jand∑jxjy*jare sufficient.b. Straightforward substitution.12.21 Sinceddπlog(π/(1-π)) = 1/(π(1-π)),Var logˆπ1-ˆπ≈1π(1-π)2π(1-π)n=1nπ(1-π)12.23 a. If∑ai= 0,EXiaiYi=Xiai[α+βxi+μ(1-δ)] =βXiaixi=βforai=xi-¯x.b.E(¯Y-β¯x) =1nXi[α+βxi+μ(1-δ)]-β¯x=α+μ(1-δ),so the least squares estimateais unbiased in the modelYi=α+βxi+i, where...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online