finding power in 1 sample t test
- From: woodland <woodland_56@xxxxxxxxx>
- Date: Thu, 23 Apr 2009 16:30:07 EDT
Here's the problem. In the null hypothesis, we have to test
Ho: mu_o = 0.82. alpha=0.05 , n=15, s=std dev =0.02456
If the true mean mu exceeds 0.82 by as much as 0.02,
is the sample size n=15 adequate to ensure that Ho: mu_o=0.82 will be rejected with probability at least 0.8 ?
In other words we have to check if the given sample size is
sufficient so that the power for this 1 sample t test is
0.8. Now for futher calculation, check the attached files.
As I have done calculation using R software, I got the power to be 0.8344 . So we can say that the sample size is adequate. But my book (Montgomery,Runger) uses MiniTab
and comes up with power=0.9117. So I want to know if
I have done something wrong.
Regards
Wood
.
- Follow-Ups:
- Re: finding power in 1 sample t test
- From: Steve Oakley
- Re: finding power in 1 sample t test
- From: Jack Tomsky
- Re: finding power in 1 sample t test
- Prev by Date: Re: data scaling and validation for learning classifier
- Next by Date: Re: finding power in 1 sample t test
- Previous by thread: Hierarchical clustrer analysis scipt
- Next by thread: Re: finding power in 1 sample t test
- Index(es):
Relevant Pages
|
