Ted this was very helpful.
To be crystal clear I'll rephrase you.
What I am interested in is the probability that my 'T' is in the tail. For
a two sided test I do not care which tail. Since (all things being equal)
the distribution of T is around 0 if the my T is greater than 0 I compare it
to
On 16-Jun-10 22:30:39, Worik R wrote:
> I have two pairs of related vectors
> x1,y1
> and
> x2,y2
>
> I wish to do a test for differences in means of x1 and y1,
> ditto x2 and y2.
>
> I am getting odd results. I am not sure I am using 'pt' properly...
> I have not included the raw vectors a
Hi Worik,
You can try
2*pt(abs(t1$statistic), t1$parameter, lower.tail=FALSE)
If the test is two sided.
Cheers,
Oscar
Oscar M. Rueda, PhD
Postdoc, Breast Cancer Functional Genomics
Cancer Research UK Cambridge Research Institute
Li Ka Shing Centre
Robinson Way
Cambridge CB2 0RE
England
O
On Jun 17, 2010, at 12:51 AM, Worik R wrote:
If it were not for the fact that I get inconsistent results I would
be sure
that I need...
I have not yet seen a proper challenge to your original assumption
that the t-statistic should be the same as the p-value. They go in
_opposite_ direct
> Sorry, I realized that is is fairly easy to test that it is an issue
> with which tail of the distribution you use. This should show what is
> going on better than my prior message.
>
> 1.353946/2 = 0.676973
> 1 - 0.676973 = 0.323027
> 0.323027 * 2 = 0.646054
>
> in pt(), the default is lower.ta
On Wed, Jun 16, 2010 at 3:30 PM, Worik R wrote:
> I have two pairs of related vectors
> x1,y1
>
> and
>
> x2,y2
>
> I wish to do a test for differences in means of x1 and y1, ditto x2 and y2.
>
> I am getting odd results. I am not sure I am using 'pt' properly...
>
> I have not included the raw v
Look what happens when you use lower.tail=FALSE (see ?pt ). If that
is not your problem, perhaps you can provide a short sample vector
that replicates your results?
Josh
On Wed, Jun 16, 2010 at 10:02 PM, Worik R wrote:
> More:
>
> When the t-stat is > 0 should I use 'pt' differently?
>
> I have
More:
When the t-stat is > 0 should I use 'pt' differently?
I have been checking my results and (except for the example I posted) all
the inconsistencies occur when t>0
Worik
On Thu, Jun 17, 2010 at 10:30 AM, Worik R wrote:
> I have two pairs of related vectors
> x1,y1
>
> and
>
> x2,y2
>
> I
If it were not for the fact that I get inconsistent results I would be sure
that I need...
2*pt(stat, df)
Section 8.1 of R-intro.pdf is explicit.
Problem is it gives inconsistent results
Worik
On Thu, Jun 17, 2010 at 10:30 AM, Worik R wrote:
> I have two pairs of related vectors
> x1,y1
>
>
9 matches
Mail list logo
