Significance of trend

2 messages · Muhammad Rahiz, Uwe Ligges

Original

Tue, Nov 1, 2011 1:47 AM #

Thanks for those who replied.

I know what a p-value is and the links given reaffirm my understanding. 
The code below gives p=0.26. This is more than 0.05 - hence null 
hypothesis i.e. attributed by chance.

But I did not specify the significance level at 0.05. So I'm wondering if 
any part of the code states so. Or does R define significance level at 
0.05 by default?

Thanks

Muhammad

On Mon, 31 Oct 2011, R. Michael Weylandt wrote:

> I'm thinking you need to read up on what a p-value is:
> https://secure.wikimedia.org/wikipedia/en/wiki/P-value
>
> Michael Weylandt
>
> On Mon, Oct 31, 2011 at 2:51 PM, Muhammad Rahiz
> <muhammad.rahiz at ouce.ox.ac.uk> wrote:
>> Hi everyone,
>>
>> I'm trying to determine the significance of a trendline. From my internet
>> search months ago, I came across the following post. I modified tim and
>> dat for simiplicity.
>>
>> tim <- 1:10
>> dat <- c(0.17, 1.09 ,0.11, 0.82, 0.23, 0.38 ,2.47 ,0.41 ,0.75, 1.44)
>>
>> fstat <- summary(lm(dat~tim))$fstatistic
>> p.val <- 1-pf(fstat["value"],fstat["numdf"],fstat["dendf"])
>>
>> The resultant p.value is 0.261. But I want to know if this
>> value is significant and at what significance level. How do I check for
>> it?
>>
>> Thanks
>>
>> --
>> Muhammad
>>
>> ______________________________________________
>> R-help at 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.
>>
>

Uwe Ligges

Tue, Nov 1, 2011 5:34 AM #

On 01.11.2011 09:47, Muhammad Rahiz wrote:

That's why you have been told to read what a p-value is! It is 
independent of the level of significance (which is not true for critical 
values). Just compare.

Uwe Ligges