[R-sig-ME] P value value for a large number of degree of freedom in lmer

Petar Milin pmilin at ff.uns.ac.rs
Tue Nov 23 20:59:18 CET 2010


On 23/11/10 20:45, Joshua Wiley wrote:
> Having a large amount of data *is* exactly what increases confidence
> in results.  A p-value is the probability of obtaining your results
> given the null hypothesis is true *in the population*.  If you have a
> lot of data, you have a lot of the population, and can more
> confidently say "this is what the population is or is note like".  The
> p-value is serving its purpose exactly as it was meant to, there is no
> need to "correct" or "alter" it.  The real question is, does anyone
> care about your effect?  Effect sizes are often a good way to get at
> the idea of is the effect meaningful, does it have practical
> significance, could an average person notice the difference?
This is very good! In some oldish statistical books you can find the 
expression "effect's substantiality".
Of course, not only the size of the sample matters. Think of having very 
large sample of clinically ill people. Would you like to generalize your 
finding to all humans, ever? Probably not, I believe. What I am trying 
to say is that not all issues can be resolved statistically. You also 
have many methodological aspects, some of which are related to the 
sampling issues.


More information about the R-sig-mixed-models mailing list