29.07.2014 Views

Asymptotic Methods in Statistical Inference - Statistics Centre

Asymptotic Methods in Statistical Inference - Statistics Centre

Asymptotic Methods in Statistical Inference - Statistics Centre

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

125<br />

at = 1. In this latter case<br />

Ã<br />

|=1 2 1<br />

= + 1 +<br />

¯<br />

1 1<br />

= +<br />

¯<br />

¯ ¯<br />

with<br />

√<br />

(ˆ − 1)<br />

!<br />

Ã<br />

1<br />

¯ + 1<br />

¯ ¯<br />

1<br />

=<br />

¯ <br />

¯<br />

!<br />

ˆ<br />

= √ <br />

Ã<br />

ˆ ˆ ¯ ¯ − ˆ ¯ ˆ ¯<br />

ˆ ¯ ˆ ¯<br />

! qˆˆ ¯ˆ ˆ ¯ <br />

The usual test, with the same asymptotic level, is obta<strong>in</strong>ed<br />

by not<strong>in</strong>g that, under , bothˆ ¯ ˆ ¯ and<br />

ˆ ˆ ¯ˆ <br />

ˆ ¯<br />

→ ¯ ¯ . Thus the denom<strong>in</strong>ator<br />

can be replaced by ˆ ˆ ¯ˆ ˆ ¯<br />

; the test statistic is<br />

then<br />

√<br />

<br />

³ˆ ˆ ¯ ¯ − ˆ ¯ ˆ ¯´<br />

=<br />

qˆˆ ¯ˆ ˆ ¯<br />

<br />

which → (0 1). For the alternative 1therejection<br />

region is ; for a two-sided alternative<br />

one can equivalently reject for 2 2 1 ().

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!