[R-sig-eco] Stepwise multiple linear regression analysis??

[Gavin Simpson]

On Tue, 2010-12-28 at 19:52 +0800, Yong Zhang wrote:
> Hi all,
> 
> Appologize for any inconvience that I brought to you in advance.
> 
> Now, I have 15 samples with species and environment data, the number
> of the environmental variable is 21. When I try to conduct the CCA in
> PC-ORD, the program stoped immediately, and I was notified that " the
> number of environmental factors should be less than the samples.What I
> would like to do is exclude some of the environmental variables. 
> 
> Could anyone tell me is the stepwise multiple linear regression
> analysis is proper for my objective?? any hint or suggestion will be
> greatly appreciated. 

This is not a general list for ecological data analysis. If you want
that, try ORDNEWS. Also, if you have a problem with PC-ORD, you can ask
the developers of that package for technical help with it.

You can't fit CCA where the number of predictors exceeds the number of
samples. With 15 predictors you would have CA which isn't constrained at
all, so again of limited use if you want to see how well the predictor
variables explain variance in the species data.

Instead of getting the computer to do your thinking for you, how about
you try to identify a smaller set of predictor variables to use with CCA
using your knowledge of the ecology of the species and the environment
in which you are working? Then fit a small set of models the are
consistent with your hypotheses about how the species and environment
relate.

Stepwise model selection is an aberration and should be avoided if at
all possible. If you must use it, learn about its issues and then avoid
it.

G

> All the best
> 
>  
> 
> 2010-12-28 
> 
> 
> 
>   
> Yong Zhang, Ph.D.
> Lab of aquatic insects & stream ecology
> Dept.of Entonology, Nanjing Agricultural University
> Nanjing, 210095,China 
> Phone number:  (+86) -25-84395241
> E-mail:2010202035 at njau.edu.cn
> 
> 
> 
> 发件人： r-sig-ecology-request 
> 发送时间： 2010-12-28  18:58:05 
> 收件人： r-sig-ecology 
> 抄送： 
> 主题： R-sig-ecology Digest, Vol 33, Issue 20 
>  
> Send R-sig-ecology mailing list submissions to
> r-sig-ecology at r-project.org
> To subscribe or unsubscribe via the World Wide Web, visit
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
> or, via email, send a message with subject or body 'help' to
> r-sig-ecology-request at r-project.org
> You can reach the person managing the list at
> r-sig-ecology-owner at r-project.org
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of R-sig-ecology digest..."
> Today's Topics:
>    1. Which analysis should I take? (Yong Zhang)
>    2. Re: Which analysis should I take? (Eduard Sz?cs)
> ----------------------------------------------------------------------
> Message: 1
> Date: Tue, 28 Dec 2010 13:53:15 +0800
> From: "Yong Zhang" <2010202035 at njau.edu.cn>
> To: "r-sig-ecology" <r-sig-ecology at r-project.org>
> Subject: [R-sig-eco] Which analysis should I take?
> Message-ID: <201012281353150154522 at njau.edu.cn>
> Content-Type: text/plain; charset="gb2312"
> Hello all, 
> Thing is like that now I have 14 sampling sites with the species and environment data. Which analysis is the best choice for determining the difference/dissimilarity among those sites. I have installed the R (2.12.2 version), which soft package or function should I take to achieve my objective.
> Thanks very much for your time and suggestion.
> Best Regards
> Yong
> 2010-12-28 
>   
> Yong Zhang, Ph.D.
> Lab of aquatic insects & stream ecology
> Dept.of Entonology, Nanjing Agricultural University
> Nanjing, 210095,China 
> Phone number:  (+86) -25-84395241
> E-mail:2010202035 at njau.edu.cn
> ------------------------------
> Message: 2
> Date: Tue, 28 Dec 2010 10:16:36 +0100
> From: Eduard Sz?cs <szoe8822 at uni-landau.de>
> To: r-sig-ecology at r-project.org
> Subject: Re: [R-sig-eco] Which analysis should I take?
> Message-ID: <4D19AAF4.2050900 at uni-landau.de>
> Content-Type: text/plain
> Hai Yong,
> have a look at the vegan package: with the function vegdist() you can 
> calculate dissimilarity indices.
> There are also some good tutorials on vegan available on the net.
> HTH,
> Eduard
> Am 28.12.2010 06:53, schrieb Yong Zhang:
> > Hello all,
> >
> > Thing is like that now I have 14 sampling sites with the species and environment data. Which analysis is the best choice for determining the difference/dissimilarity among those sites. I have installed the R (2.12.2 version), which soft package or function should I take to achieve my objective.
> >
> > Thanks very much for your time and suggestion.
> >
> > Best Regards
> >
> > Yong
> > 2010-12-28
> >
> >
> >
> >
> > Yong Zhang, Ph.D.
> > Lab of aquatic insects&  stream ecology
> > Dept.of Entonology, Nanjing Agricultural University
> > Nanjing, 210095,China
> > Phone number:  (+86) -25-84395241
> > E-mail:2010202035 at njau.edu.cn
> >
> >
> > _______________________________________________
> > R-sig-ecology mailing list
> > R-sig-ecology at r-project.org
> > https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
> [[alternative HTML version deleted]]
> ------------------------------
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
> End of R-sig-ecology Digest, Vol 33, Issue 20
> *********************************************
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology

-- 
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%
 Dr. Gavin Simpson             [t] +44 (0)20 7679 0522
 ECRC, UCL Geography,          [f] +44 (0)20 7679 0565
 Pearson Building,             [e] gavin.simpsonATNOSPAMucl.ac.uk
 Gower Street, London          [w] http://www.ucl.ac.uk/~ucfagls/
 UK. WC1E 6BT.                 [w] http://www.freshwaters.org.uk
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%