[R] Analyzing dissimilarity ratings with Multidimensional Scaling

Tue Feb 15 13:39:21 CET 2011

What was wrong with the first example on google with distance
between cities where I used the daisy output for a distance matrix?
In all the excitement I lost track of what your rows and columns
are but presumably you now have the heatmap example
with dendogram, daisy for distances, and an mds example.
What pieces are missing?

________________________________
> Date: Tue, 15 Feb 2011 13:35:33 +0100
> Subject: Re: [R] Analyzing dissimilarity ratings with Multidimensional
> Scaling
> From: turchet.luca at gmail.com
> To: djmuser at gmail.com
> CC: marchywka at hotmail.com; r-help at r-project.org
>
> Hi Dennis and Mike,
> thanks for the links, but of course I already knew them.
>
> So far I understood that likely the MDS is not the right analysis,
> since my table
> does not contain distances. Indeed if you have a look to the diagonal of my
> table you can see that the values are not 0 (which should express zero
> distance
> between the same stimuli).
> The values I have in my table are not distances but average scores from
> a 9 points
> Likert scale, assessing the degree of coherence between two modalities
> (audio and
> visual) simulating different pairs of materials:
>
>
> WD MT SW GR SN DL
> WD 7.00 6.50 4.91 4.83 5.50 5.00
> MT 7.33 6.91 2.08 3.16 4.25 3.25
> SW 2.91 1.75 7.91 6.25 6.83 5.41
> GR 2.91 2.66 6.25 6.41 7.25 6.75
> SN 4.00 4.00 5.58 6.00 7.00 6.58
> DL 3.91 3.08 5.16 6.25 6.50 6.83
>
>
> So my question now is: which analysis should I perform in your opinion?
>
>
> Thanks in advance
>
> Luca
>
>
>
> On Tue, Feb 15, 2011 at 11:01 AM, Dennis Murphy
> > wrote:
> Hi:
>
> Try the following for examples of the use of MDS in R:
>
> http://www.statmethods.net/advstats/mds.html
> http://cran.r-project.org/web/packages/HSAUR/vignettes/Ch_multidimensional_scaling.pdf
> http://cran.r-project.org/web/packages/HSAUR2/vignettes/Ch_multidimensional_scaling.pdf
> # (2nd ed.)
>
> http://www.personality-project.org/R/mds.html
> http://cran.r-project.org/web/packages/HSAUR2/vignettes/Ch_multidimensional_scaling.pdf
> http://polisci.msu.edu/jacoby/icpsr/scaling/handouts/mds3/Nonmetric%20MDS,%202009%20Handout.pdf
>
> Gotten by Googling 'multidimensional scaling R'...
>
> HTH,
> Dennis
>
> On Mon, Feb 14, 2011 at 3:53 PM, Luca Turchet
> > wrote:
> Hi Mike and all,
> thanks again. Actually your suggestion is a good way to print data in R, but
> unfortunately
> does not add any information to what I aready have on my table data.
>
> What I would like to see is how far are the trials from each others in terms
> of
> similarity.
>
> ...maybe I just need another way to analyze the data. For this reason I ask
> if someone has other suggestions, maybe coming back to the MDS?
>
> Is there anyone who can provide an R example about how I can analyze my data
> with a MDS analysis?
>
>
> Luca
>
>
>
> On Mon, Feb 14, 2011 at 6:40 PM, Mike Marchywka
> >wrote:
>
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > ________________________________
> > > Date: Mon, 14 Feb 2011 17:55:50 +0100
> > > Subject: Re: [R] Analyzing dissimilarity ratings with Multidimensional
> > > Scaling
> > > From: turchet.luca at gmail.com
> > > To: marchywka at hotmail.com
> > > CC: r-help at r-project.org
> > >
> > > Dear Mike,
> > > thanks a lot for your answer. Unfortunately the way you kindly
> > > suggested is not suitable
> > > to solve my problem.
> > >
> > > Indeed as said, I need to display the distances between the proposed
> > trials,
> > > for example I would like to see a 2D plot where I can see how fare is the
> > > trial MT-MT from the trial MT-SW (that is how far are the evaluations of
> > the
> > > audio-visual trial with metal-metal from the trial with metal-snow)
> >
> > I thought the dendrogram is what you wanted but apparently you can just
> > use daisy perhaps. This had a lot of typos and may not run as last example
> > I sent had bad edits but this does create scatterplot with zero points on
> > diagonal etc. Presumably you can define your own functions appropriate for
> > your data ( I've never used some of this before which is why I'm trying,
> > cavear emptor).
> >
> > library("cluster")
> > ddm<-daisy(matrix(xm,6,6),metric="euclidean")
> > str(ddm)
> > library("scatterplot3d")
> > dasm<-as.matrix(ddm);
> > nddf<-reshape(data.frame(dasm),varying=c("X1","X2","X3","X4","X5","X6"),
> > v.name="time", direction="long")
> > str(nddf)
> > ij<-as.numeric(gl(6,6))
> > n2<-cbind(nddf,ij)
> > str(n2)
> > scatterplot3d(n2$ij,n2$id,n2$time,type="h")
> >
> >
> >
> >
> > >
> > > Does anyone have any other suggestion to build such perceptual map in R?
> > >
> > > In addition, how can I get some information regarding the significant
> > > differences
> > > between the trials?
> > > If I used the ANOVA would be simple to get those p-values. I wonder if
> > > it is the
> > > case to use it...
> > >
> > > Help!
> > >
> > >
> > > Luca
> > >
> > >
> > >
> > >
> > > On Mon, Feb 14, 2011 at 3:23 PM, Mike Marchywka
> > > > wrote:
> > >
> > >
> > >
> > > > My first goal in the analysis process is to print a perceptual map
> > where to
> > > > place the pairs of
> > > > audio-visual stimuli (e.g. WD-WD, MT-DL, etc.) and see how far the
> > trials
> > > > are from each other.
> > >
> > > I've been using heatmap for stuff like this.
> > > You can get a nice picture this way and get quick visual
> > > survey and dendrograms,
> > >
> > > xm<-scan(file="avm.txt")
> > > str(xm)
> > > heatmap(xm)
> > > heatmap(matrix(xm,6,6))
> > >
> > > I ran the above on your data and it visually looks
> > > like there could be interesting patterns to test.
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > --
> > > ---------------------------------------------------------------
> > >
> > > "Music is a moral law:
> > > It gives a soul to the Universe,
> > > wings to the mind,
> > > flight to the imagination,
> > > a charm to sadness,
> > > and life to everything.
> > > It is the essence of order,
> > > and leads to all that is good,
> > > just and beautiful,
> > > of which it is the invisible,
> > > but nevertheless dazzling,
> > > passionate, and eternal form".
> > >
> > > Plato, 400 B.C. (from the Dialogues)
> > >
> >
> >
>
>
>
> --
> ---------------------------------------------------------------
>
> "Music is a moral law:
> It gives a soul to the Universe,
> wings to the mind,
> flight to the imagination,
> a charm to sadness,
> and life to everything.
> It is the essence of order,
> and leads to all that is good,
> just and beautiful,
> of which it is the invisible,
> but nevertheless dazzling,
> passionate, and eternal form".
>
> Plato, 400 B.C. (from the Dialogues)
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at 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.
>
>
>
>
> --
> ---------------------------------------------------------------
>
> "Music is a moral law:
> It gives a soul to the Universe,
> wings to the mind,
> flight to the imagination,
> a charm to sadness,
> and life to everything.
> It is the essence of order,
> and leads to all that is good,
> just and beautiful,
> of which it is the invisible,
> but nevertheless dazzling,
> passionate, and eternal form".
>
> Plato, 400 B.C. (from the Dialogues)
>