Re: [R] ttest in R

2009-03-08 Thread Thomas Petzoldt
1Rnwb wrote: since the estimation is not done pairwise so i cannot use pairwise.t.test, how do i apply tukeyHSD Note correct capitalization: TukeyHSD and follow the examples on the help page: ?TukeyHSD You may also contact a statistics textbook. In addition I send you some links off list.

Re: [R] ttest in R

2009-03-08 Thread 1Rnwb
since the estimation is not done pairwise so i cannot use pairwise.t.test, how do i apply tukeyHSD Thomas Petzoldt-4 wrote: > > 1Rnwb schrieb: >> Thanks for the help, but ANOVA will give me a single pvalue, then how i >> can >> make sure which group is showing the significant differences. > > H

Re: [R] ttest in R

2009-03-08 Thread Thomas Petzoldt
1Rnwb schrieb: Thanks for the help, but ANOVA will give me a single pvalue, then how i can make sure which group is showing the significant differences. Hi, ANOVA is fine and please have a look on ?TukeyHSD and don't forget to consult your statistics textbook about post-hoc testing. If you in

Re: [R] ttest in R

2009-03-08 Thread 1Rnwb
Thanks for the help, but ANOVA will give me a single pvalue, then how i can make sure which group is showing the significant differences. Tal Galili wrote: > > Hello friend. > > I believe anova might be a better solution for you. > You might have a look here: > http://www.personality-project.or

Re: [R] ttest in R

2009-03-07 Thread Tal Galili
p.s: since the "data" (Y variable) is very much not normal (also after a log transform)I would consider going with an a-parametric test check: ? wilcox.test (for a non parametric t.test ) OR (for a non parametric simple anova) ?kruskal.test On Sat, Mar 7, 2009 at 11:51 PM, Tal Galili wrote: >

Re: [R] ttest in R

2009-03-07 Thread Tal Galili
Hello friend. I believe anova might be a better solution for you. You might have a look here: http://www.personality-project.org/r/r.anova.html A simple R session that will work for you is: # getting the data in: data1 <- read.table( "enter the path of the file here, look at ?read.table for exa

[R] ttest in R

2009-03-07 Thread 1Rnwb
Dear list, i am a biologist who needs to do some ttest between disease and non disease, sex, genotype and the serum levels of proteins on a large number of individuals. i have been using excel for a long time but it is very tedious and time consuming. i am posting the data below and ask your hel