[R] pairwise.t.test for paired data

[Peter Dalgaard]

francogrex wrote:
> Dear R-help, I have a question about pairwise.t.test and adjustment for
> multiple comparisons for paired data points.
> I have the following data:
>
> n=c("x", "x", "x", "x", "x", "x", "x", "x", "x", "x", "y", "y", 
> "y", "y", "y", "y", "y", "y", "y", "y", "z", "z", "z", "z", "z", 
> "z", "z", "z", "z", "z")
>
> and 
>
> d=c(3.38501220258169, 0.732526675324789, -0.638736410843593,
> 7.87414894080545, 
> 3.6341475242877, -2.80762187242111, 14.1405912463741, -9.29904247482419, 
> 9.8636915453275, 5.43287495347734, 12.2804612051493, 7.83975161028828, 
> 15.2200332220459, 6.72953902619773, 9.11557944514228, 9.79268012944973, 
> 7.6112143261283, 10.3912598689021, 8.04934992955112, 13.3935024789408, 
> 12.1276080143311, 11.2987640203845, 2.05386456839974, 12.8386335687123, 
> 3.84418682233211, 17.4626505871045, 6.5947625121599, -10.6664750811034, 
> 21.3346023277454, 4.99216977422232)
>
> Then I do the tests
>
>   
>> pairwise.t.test(d,n,p.adjust="bonf",paired=F)
>>     
>
>         Pairwise comparisons using t tests with pooled SD 
>
> data:  d and n 
>
>   x     y    
> y 0.092 -    
> z 0.326 1.000
>
> P value adjustment method: bonferroni
>
> Assuming the data are paired:
>
>   
>> pairwise.t.test(d,n,p.adjust="bonf",paired=T)
>>     
>
>         Pairwise comparisons using t tests with pooled SD 
>
> data:  d and n 
>
>   x     y    
> y 0.092 -    
> z 0.326 1.000
>
> P value adjustment method: bonferroni 
>
> I have no indication that the test is taking into account that the data are
> paired! Is it possible to do a pairwise t test with adjustement for multiple
> comparisons using pairwise.t.test? If not any other function for that?
> Thanks
Get rid of the pooled SD, and pairwise.t.test() will actually do the 
paired tests, although that is not what the function was designed to do. 
(Oddly enough, this issue hasn't come up in the 6+ years that the 
function has existed, and then it pops up twice with little over a week 
between, see the thread started by James Root on March 26.)

-- 
   O__  ---- Peter Dalgaard             Øster Farimagsgade 5, Entr.B
  c/ /'_ --- Dept. of Biostatistics     PO Box 2099, 1014 Cph. K
 (*) \(*) -- University of Copenhagen   Denmark      Ph:  (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)              FAX: (+45) 35327907