Hi,
If you mean if the t.test is done as if the samples where paired, the answer is
yes if you write argument paired = TRUE; and the pairs are done in order, that
is 1º with 1º, 2º with 2º, etc.
As you wrote, (paired = FALSE) the t.test is unpaired, and the order of elements
in the vectors are not taken into account.
Regards
Patricia
> Date: Thu, 29 Jan 2009 13:59:41 +0000
> From: amitrhelp@yahoo.co.uk
> To: r-help@r-project.org
> Subject: [R] t-test
>
> When doing the t-test in the below manner will r compare each element of
the array with the relevant one. I.e. if i was comparing x and y would (1 and 0)
and (1 and 9) be treated as separate variables. Or does it just assume one
variable.
>
>
> # test data
>
> x <- c(1,1.1,1.15,1.2,1.21,1.23)
>
> y <- c(0.9,1,1.16,1.18,1.19,1.2)
>
> z <- c(1.4,1.42,1.43,1.44,1.45,1.46)
>
>
> ###Â Student's t-test
>
> # for help in R type ?t.test()
>
> # defaults are:
> # alternative = "two.sided" i.e. two-sided t-test
> # var.equal = FALSE i.e. unequal variance
>
> # note:
> # na.rm = TRUE removes missing values
> # $p.value gives the p-value for the test
>
> t.test(x, y, na.rm=TRUE, paired=FALSE)$p.value
>
> # gives 0.5026467
>
> t.test(x, z, na.rm=TRUE, paired=FALSE)$p.value
>
> # gives 0.0003166352
>
>
>
>
> [[alternative HTML version deleted]]
>
_________________________________________________________________
[[alternative HTML version deleted]]