On May 8, 2010, at 9:38 PM, Duncan Murdoch wrote:
On 08/05/2010 9:14 PM, Joris Meys wrote:
On Sat, May 8, 2010 at 7:02 PM, Bak Kuss <bakk...@gmail.com> wrote:
Just wondering.
The smallest the p-value, the closer to 'reality' (the more
accurate)
the model is supposed to (not) be (?).
How realistic is it to be that (un-) real?
That's a common misconception. A p-value expresses no more than the
chance
of obtaining the dataset you observe, given that your null
hypothesis _and
your assumptions_ are true.
I'd say it expresses even less than that. A p-value is simply a
transformation of the test statistic to a standard scale. In the
nicer situations, if the null hypothesis is true, it'll have a
uniform distribution on [0,1]. If H0 is false but the truth lies in
the direction of the alternative hypothesis, the p-value should have
a distribution that usually gives smaller values. So an unusually
small value is a sign that H0 is false: you don't see values like
1e-6 from a U(0,1) distribution very often, but that could be a
common outcome under the alternative hypothesis. (The not so nice
situations make things a bit more complicated, because the p-value
might have a discrete distribution, or a distribution that tends
towards large values, or the U(0,1) null distribution might be a
limiting approximation.)
So to answer Bak, the answer is that yes, a well-designed statistic
will give p-values that tend to be smaller the further the true
model gets from the hypothesized one, i.e. smaller p-values are
probably associated with larger departures from the null. But the p-
value is not a good way to estimate that distance. Use a parameter
estimate instead.
And. Thank you for this paper. As a non-statistician I found it most
instructive:
http://pubs.amstat.org/doi/pdfplus/10.1198/000313008X332421
--
David.
Duncan Murdoch
Essentially, a p-value is as "real" as your
assumptions. In that way I can understand what Robert wants to say.
But with
lare enough datasets, bootstrapping or permutation tests gives
often about
the same p-value as the asymptotic approximation. At that moment, the
central limit theorem comes into play, which says that when the
sample size
is big enough, the mean is -close to- normally distributed. In
those cases,
the test statistic also follows the proposed distribution and your
p-value
is closer to "reality". Mind you, the "sample size" for a specific
statistic
is not always merely the number of observations, especially in more
advanced
methods. Plus, violations of other assumptions, like independence
of the
observations, changes the picture again.
The point is : what is reality? As Duncan said, a small p-value
indicates
that your null hypothesis is not true. That's exactly what you look
for,
because that is the proof the relation in your dataset you're
looking at,
did not emerge merely by chance. You're not out to calculate the
exact
chance. Robert is right, reporting an exact p-value of 1.23 e-7
doesn't make
sense at all. But the rejection of your null-hypothesis is as real
as life.
The trick is to test the correct null hypothesis, and that's were
it most
often goes wrong...
Cheers
Joris
bak
p.s. I am no statistician
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
David Winsemius, MD
West Hartford, CT
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.