Hi:
It strikes me as a little curious that the No Virus values in each of your
example data sets
are all *exactly* 1. Why is that?
Dennis
On Thu, Apr 29, 2010 at 4:52 AM, Yanwei Tan
<Tan@nbio.uni-heidelberg.de>wrote:
> Dear all,
>
> I have a quite basic questions about anova analysis in R, sorry for this,
> but I have no clue how to explain this result.
>
> I have two datasets which are named: nmda123, nmda456. Each dataset has
> three samples which were measured three times. And I would like to compare
> means of them with Posthoc test using R, following please see the output:
>
> (CREB, mCREB and No virus are the name of samples)
>
> > nmda123
> Values ind
> 1 6.7171265 CREB
> 2 5.0343117 CREB
> 3 6.9000000 CREB
> 4 0.1195394 mCREB
> 5 0.1221876 mCREB
> 6 0.1900000 mCREB
> 7 1.0000000 No Virus
> 8 1.0000000 No Virus
> 9 1.0000000 No Virus
>
> > nmda456
> Values ind
> 1 6.4486940 CREB
> 2 6.2277490 CREB
> 3 6.5000000 CREB
> 4 0.2000000 mCREB
> 5 0.3766052 mCREB
> 6 0.4000000 mCREB
> 7 1.0000000 No Virus
> 8 1.0000000 No Virus
> 9 1.0000000 No Virus
>
> > TukeyHSD(aov(Values ~ ind, data = nmda456))
> Tukey multiple comparisons of means
> 95% family-wise confidence level
>
> Fit: aov(formula = Values ~ ind, data = nmda456)
>
> $ind
> diff lwr upr p adj
> mCREB-CREB -6.0666126 -6.3289033 -5.8043219 0.0000000
> No Virus-CREB -5.3921477 -5.6544383 -5.1298570 0.0000000
> No Virus-mCREB 0.6744649 0.4121743 0.9367556 0.0005382
>
> > TukeyHSD(aov(Values ~ ind, data = nmda123))
> Tukey multiple comparisons of means
> 95% family-wise confidence level
>
> Fit: aov(formula = Values ~ ind, data = nmda123)
>
> $ind
> diff lwr upr p adj
> mCREB-CREB -6.073237 -7.5618886 -4.584585 0.0000392
> No Virus-CREB -5.217146 -6.7057976 -3.728495 0.0000943
> No Virus-mCREB 0.856091 -0.6325606 2.344743 0.2588450
>
> So my question is No virus-mCREB group. Even I looked at the data by eyes,
> there is big difference between no virus and mCREB in data nmda123, but why
> the pvalue is 0.2588450, but in nmda456 data, the pvalue is 0.0005382. But
> I can see there is bigger difference in nmda123 than nmda456, I do not
know
> why. Sorry for my inexperiences in statistics.
>
> Thanks for your reply and time!
>
> Cheers,
> Wei
>
> ______________________________________________
> R-help@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.
>
[[alternative HTML version deleted]]