sol9 - ' % (Q D % A ) ' i B(C&P98 E @ ' ' ' f % Phgy(fQ...

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

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: ' % (Q D % A ) ' i B(C&P98 E @ ' ' ' f % Phgy(fQ F E 0 ) @ exP E 5 63 D A @ f 8 (da(B ' @ (& D ' A ' @ f 8 aay(&T 0 ) @ VP u @ @ ( ) @ ) P& b X E E u E E cW X f X D YW x8 YW D b 8 cW f b b u YW D bcW f8 E yW b DX b 8 s q p u D cW (rvD . To compute the X fE X xw8 YW X 8 s q YW a(rp so E E b X 8 s q tD cW f YW (rp E X @ @ @ I aP` YW (2) standard error: and E ) @ ) f @ I PihP&Qge F E b @ @ ) (Pd cW E D A ' 8 5 3 ) ' % P(CBVUTSR4(&Q0 F ) ' % (& (1) and . For the confidence interval, and . For the median, . Everyone will get a different bootstrap standard error. I get . The normal interval is . For the percentile interval I get . # ! $" DCBVU8 A ' E ) I E P&H % (G F D A @ (CB(98 5 ) ' % 0 764(&21 3 Hence, E E b X b cW f YW (rp 63 X 8 s q 5 b f b u D YW x8 W D X f X D yW Sw8 cW An 80 per cent confidence interval is confidence interval is E @ @ I & E 5 R3 (3) cent confidence interval In this case we get and the estimated standard error is Homework 9 Solutions . From the bootstrap I get . E (4) We may represent the bootstrap distribution as: 1 . . A 95 per cent and a 95 per D aVrp 8 s q " f D a(rp 8 s q " # ! f b b u b ef b b u D b U8 ef b D (rvu b 8 s q p D b ef 6A aaaA X 8 (rvD RA aa A X (rp s q p u 8 s q b b ef b f D 6A aaaA X (rp 8 s q b where D 8 (rp s q D 6A aaA X a(rp 8 s q DX 8 where . Now D 8 so D 6A a aA X 8 . . . D 6A aaaA X 8 . . . X X X D 8 b X and D 8 rD 6A aaA X 8 D 8 and D ef y b D f 8 8 D 6A aaA X a(rp 8 s q b D f 8 . Hence, Also, Treat the 's as fixed. Now 2 $ HD 8 (rp s q ...
View Full Document

This note was uploaded on 05/25/2008 for the course STAT 36625 taught by Professor Larrywasserman during the Fall '01 term at Carnegie Mellon.

Ask a homework question - tutors are online