{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

slides_13_inferasymptotic

# In fact this convergence holds if we replace 165 by

This preview shows pages 24–33. Sign up to view the full content.

(In fact, this convergence holds if we replace 1.65 by any critical value c .) 24

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
We have shown that the asymptotic power function for this example is 0, 0 0 .05 1, 0 Unfortunately, this is unhelpful for determining what the likely power is for a given sample size. Most sensible tests for this problem will have exactly the same asymptotic power function. In fact, if one uses only the even-numbers observations (so only half of the data), the asymptotic power function is the same. 25
In the case of a normal population, we computed the exact power function for any n , and it does not look much like the asymptotic power function (although we can see it is converging to the asymptotic power function). 26

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
0 .2 .5 1 -1 -.5 0 .5 1 mu exact power asymptotic power Exact Power (n = 25) and Asymptotic Power 27
.05 .2 .5 1 -1 -.5 0 .5 1 mu exact power asymptotic power Exact Power (n = 100) and Asymptotic Power 28

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document