Hi all, I wondered whether anyone has some advice on a stats-related 'sanity check', as I ran a nonparametric multivariate test (mulrank function as decribed by R. Wilcox, 2005) on both systems, but got different results (please see below for the system-specific outputs)! The functions I used are attached as well. Any advice would be much appreciated! Thanks in advance for getting back to me! Best wishes, Mareen ------------------------------------ Mac:> data03<-selby2(data02, c(1,2), 3) > mulrank(3,6,data03$x)$test.stat [1] 0.9331133 $nu1 [1] 11.46300 $p.value [,1] [1,] 0.509296 $N [1] 233 $q.hat [,1] [,2] [,3] [,4] [,5] [,6] [1,] 0.4940071 0.5256726 0.5176384 0.5476290 0.4690935 0.5265100 [2,] 0.5170627 0.4791950 0.5026431 0.4867843 0.4778865 0.5033497 [3,] 0.4680729 0.4944258 0.4889563 0.4505391 0.5311420 0.4726002 Win:> mulrank(3,6, data03$x)$test.stat [1] 1.114665 $nu1 [1] 8.155991 $p.value [,1] [1,] 0.3491221 $N [1] 233 $q.hat [,1] [,2] [,3] [,4] [,5] [,6] [1,] 0.4940071 0.5406417 0.5236837 0.5656338 0.4771799 0.5324505 [2,] 0.5162776 0.4801895 0.5022244 0.4960745 0.4854234 0.4820737 [3,] 0.5013608 0.4920967 0.4810269 0.4482885 0.5326861 0.4871506 nabble.com/file/p23595008/Rallfun-v92.txt Rallfun-v92.txt -- View this message in context: nabble.com/MAC-OSX-vs-Win-XP:-Different-stats-test-results!-tp23595008p23595008.html Sent from the R help mailing list archive at Nabble.com.
Mareen wrote:> Hi all, > I wondered whether anyone has some advice on a stats-related 'sanity check', > as I ran a nonparametric multivariate test (mulrank function as decribed by > R. Wilcox, 2005) on both systems, but got different results (please see > below for the system-specific outputs)! The functions I used are attached as > well. Any advice would be much appreciated! Thanks in advance for getting > back to me!Perhaps you should show us a summary(data03$x) on both systems. It could be that you are not working with the same data... -p> > Best wishes, > Mareen > ------------------------------------ > Mac: > >> data03<-selby2(data02, c(1,2), 3) >> mulrank(3,6,data03$x) > $test.stat > [1] 0.9331133 > > $nu1 > [1] 11.46300 > > $p.value > [,1] > [1,] 0.509296 > > $N > [1] 233 > > $q.hat > [,1] [,2] [,3] [,4] [,5] [,6] > [1,] 0.4940071 0.5256726 0.5176384 0.5476290 0.4690935 0.5265100 > [2,] 0.5170627 0.4791950 0.5026431 0.4867843 0.4778865 0.5033497 > [3,] 0.4680729 0.4944258 0.4889563 0.4505391 0.5311420 0.4726002 > > Win: >> mulrank(3,6, data03$x) > $test.stat > [1] 1.114665 > > $nu1 > [1] 8.155991 > > $p.value > [,1] > [1,] 0.3491221 > > $N > [1] 233 > > $q.hat > [,1] [,2] [,3] [,4] [,5] [,6] > [1,] 0.4940071 0.5406417 0.5236837 0.5656338 0.4771799 0.5324505 > [2,] 0.5162776 0.4801895 0.5022244 0.4960745 0.4854234 0.4820737 > [3,] 0.5013608 0.4920967 0.4810269 0.4482885 0.5326861 0.4871506 > > nabble.com/file/p23595008/Rallfun-v92.txt Rallfun-v92.txt >-- O__ ---- Peter Dalgaard ?ster 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
If mulrank does any sort of random number generation or non-exhaustive randomization, you should set the seed of the random number generator first: set.seed(1) mulrank(3,6,data03$x) On Mon, May 18, 2009 at 7:37 AM, Mareen <mareenweber at yahoo.com> wrote:> > Hi all, > I wondered whether anyone has some advice on a stats-related 'sanity check', > as I ran a nonparametric multivariate test (mulrank function as decribed by > R. Wilcox, 2005) on both systems, but got different results (please see > below for the system-specific outputs)! The functions I used are attached as > well. Any advice would be much appreciated! Thanks in advance for getting > back to me! > > Best wishes, > Mareen > ------------------------------------ > Mac: > >> data03<-selby2(data02, c(1,2), 3) >> mulrank(3,6,data03$x) > $test.stat > [1] 0.9331133 > > $nu1 > [1] 11.46300 > > $p.value > ? ? ? ? [,1] > [1,] 0.509296 > > $N > [1] 233 > > $q.hat > ? ? ? ? ?[,1] ? ? ?[,2] ? ? ?[,3] ? ? ?[,4] ? ? ?[,5] ? ? ?[,6] > [1,] 0.4940071 0.5256726 0.5176384 0.5476290 0.4690935 0.5265100 > [2,] 0.5170627 0.4791950 0.5026431 0.4867843 0.4778865 0.5033497 > [3,] 0.4680729 0.4944258 0.4889563 0.4505391 0.5311420 0.4726002 > > Win: >> mulrank(3,6, data03$x) > $test.stat > [1] 1.114665 > > $nu1 > [1] 8.155991 > > $p.value > ? ? ? ? ?[,1] > [1,] 0.3491221 > > $N > [1] 233 > > $q.hat > ? ? ? ? ?[,1] ? ? ?[,2] ? ? ?[,3] ? ? ?[,4] ? ? ?[,5] ? ? ?[,6] > [1,] 0.4940071 0.5406417 0.5236837 0.5656338 0.4771799 0.5324505 > [2,] 0.5162776 0.4801895 0.5022244 0.4960745 0.4854234 0.4820737 > [3,] 0.5013608 0.4920967 0.4810269 0.4482885 0.5326861 0.4871506 > > nabble.com/file/p23595008/Rallfun-v92.txt Rallfun-v92.txt > > -- > View this message in context: nabble.com/MAC-OSX-vs-Win-XP:-Different-stats-test-results!-tp23595008p23595008.html > Sent from the R help mailing list archive at Nabble.com. > > ______________________________________________ > R-help at r-project.org mailing list > stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. >-- Mike Lawrence Graduate Student Department of Psychology Dalhousie University Looking to arrange a meeting? Check my public calendar: tr.im/mikes_public_calendar ~ Certainty is folly... I think. ~
Hello Peter, thanks for your quick response - you have hinted into the right direction. I triple-checked my data and - how embarrassing - I actually did have different data matrices. Due to a faulty Win Excel macro that sorted data into groups after data collection, I ended up with different data in my 3 groups in the Win "test" ... All good now :-). Thanks again! Best wishes, Mareen
A good example of why manipulating data in external programs is dangerous. See a discussion of this at tolstoy.newcastle.edu.au/R/e6/help/09/05/13697.html On Mon, May 18, 2009 at 10:56 AM, Mareen Weber <mareenweber at yahoo.com> wrote:> > Hello Peter, > thanks for your quick response - you have hinted into the right direction. I triple-checked my data and - how embarrassing - I actually did have different data matrices. Due to a faulty Win Excel macro that sorted data into groups after data collection, I ended up with different data in my 3 groups in the Win "test" ... All good now :-). Thanks again! > > Best wishes, > Mareen > > > > > > ______________________________________________ > R-help at r-project.org mailing list > stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. >-- Mike Lawrence Graduate Student Department of Psychology Dalhousie University Looking to arrange a meeting? Check my public calendar: tr.im/mikes_public_calendar ~ Certainty is folly... I think. ~