[R] SPSS -> R

John Fox jfox at mcmaster.ca
Wed Nov 30 16:07:28 CET 2011


Dear Michael,

The original poster and I ended up pursuing this question off-list and, I
think, resolving the issue, discovering in the process that SPSS apparently
reports "type-II" tests that are really "type-III" tests. (Yes, one fits a
multivariate linear model prior to calling Anova() with the idata and
idesign arguments.)

Best,
 John

--------------------------------
John Fox
Senator William McMaster
  Professor of Social Statistics
Department of Sociology
McMaster University
Hamilton, Ontario, Canada
http://socserv.mcmaster.ca/jfox




> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
> project.org] On Behalf Of Michael Dewey
> Sent: November-30-11 6:49 AM
> To: Kristi Shoemaker; r-help at r-project.org
> Subject: Re: [R] SPSS -> R
> 
> At 16:41 27/11/2011, Kristi Shoemaker wrote:
> >Hi John,
> >
> >Your assumptions are correct and those examples were very helpful,
> thanks!
> >
> >I think I'm almost there, but I'm screwing up something with the
> >within-subjects factor (the example has two, I only have one). See
> >below. Why am I not seeing my within-subjects factors in the ANOVA
> report?
> 
> I am by no means an expert in this but my reading of the documentation
> suggests that the idata and idesign parameters are for multivariate
> models and that is not what you specified.
> 
> 
> 
> > > Dat2
> >    Subj  Age Sex Time   HippV
> >1   s01 32.9   1   0w 6.50098
> >2   s01 32.9   1   6w 6.91793
> >3   s02 35.1   0   0w 7.32480
> >4   s02 35.1   0   6w 7.56012
> >5   s03 34.4   0   0w 6.51385
> >6   s03 34.4   0   6w 6.56875
> >9   s05 39.9   1   0w 6.92855
> >10  s05 39.9   1   6w 6.94926
> >11  s06 29.5   1   0w 6.99383
> >12  s06 29.5   1   6w 7.10568
> >13  s07 45.9   1   0w 6.94380
> >14  s07 45.9   1   6w 7.08190
> >15  s08 20.3   1   0w 7.76881
> >16  s08 20.3   1   6w 7.72725
> >17  s09 26.9   0   0w 5.37566
> >18  s09 26.9   0   6w 5.74887
> >21  s11 22.0   0   0w 7.12992
> >22  s11 22.0   0   6w 7.16237
> >23  s12 31.0   1   0w 6.70629
> >24  s12 31.0   1   6w 6.80872
> >25  s13 50.1   1   0w 7.22649
> >26  s13 50.1   1   6w 7.58900
> >27  s14 22.2   0   0w 5.97577
> >28  s14 22.2   0   6w 5.80801
> >29  s16 20.4   1   0w 7.99554
> >30  s16 20.4   1   6w 8.09260
> >31  s20 24.0   0   0w 7.01014
> >32  s20 24.0   0   6w 6.87821
> >33  s21 32.4   0   0w 5.90883
> >34  s21 32.4   0   6w 5.95392
> >37  s24 22.2   0   0w 6.15474
> >38  s24 22.2   0   6w 6.00906
> >41  s27 22.2   1   0w 7.88765
> >42  s27 22.2   1   6w 7.76038
> >49  s35 24.3   0   0w 6.05998
> >50  s35 24.3   0   6w 6.07399
> >51  s36 23.5   0   0w 7.83182
> >52  s36 23.5   0   6w 7.60268
> >53  s38 59.7   1   0w 7.39672
> >54  s38 59.7   1   6w 6.98291
> >55  s39 40.5   0   0w 7.31330
> >56  s39 40.5   0   6w 7.50559
> >57  s40 24.2   1   0w 8.54958
> >58  s40 24.2   1   6w 8.65016
> >59  s41 23.6   1   0w 7.76049
> >60  s41 23.6   1   6w 7.58946
> >61  s42 53.3   0   0w 7.03388
> >62  s42 53.3   0   6w 7.48384
> >63  s43 34.4   0   0w 6.86967
> >64  s43 34.4   0   6w 6.81076
> >65  s44 44.8   0   0w 7.33779
> >66  s44 44.8   0   6w 7.86175
> >67  s45 40.2   1   0w 6.55963
> >68  s45 40.2   1   6w 6.52577
> >71  s47 26.5   0   0w 7.09418
> >72  s47 26.5   0   6w 6.92850
> >73  s48 22.7   0   0w 6.77078
> >74  s48 22.7   0   6w 6.67289
> >75  s50 36.1   1   0w 7.47208
> >76  s50 36.1   1   6w 7.55876
> > > library(car)
> >Loading required package: MASS
> >Loading required package: nnet
> >Loading required package: survival
> >Loading required package: splines
> > > timepoints <- factor(c("0w", "6w"),
> >+                 levels=c("0w","6w"))
> > > idata=data.frame(timepoints)
> > > mod.ok <- lm(HippV ~ Age*Sex,data=Dat2) (av.ok <- Anova(mod.ok,
> > > idata=idata, idesign=~timepoints))
> >Anova Table (Type II tests)
> >
> >Response: HippV
> >            Sum Sq Df F value    Pr(>F)
> >Age        0.0234  1   0.071 0.7909202
> >Sex        5.3082  1  16.128 0.0001781 ***
> >Age:Sex    4.7772  1  14.514 0.0003478 ***
> >Residuals 18.4314 56
> >---
> >Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
> > >
> >
> >
> >
> >________________________________
> >  From: John Fox <jfox at mcmaster.ca>
> >
> >Cc: r-help at r-project.org
> >Sent: Saturday, November 26, 2011 7:19 PM
> >Subject: RE: [R] SPSS -> R
> >
> >Dear Kristi,
> >
> >I assume that this is a repeated-measures ANOVA with one
> >within-subjects factor (Time) and two between-subjects factors (Age
> and
> >Sex, which are crossed). If Age is numeric, and not a factor, then the
> >"type-III" tests that you requested don't test sensible hypotheses. In
> >any event, if my guess is right about the design, then you can use the
> >Anova() function in the car package for an equivalent analysis. See
> the
> >repeated-measures example in ?Anova (for the O'Brien and Kaiser data).
> >
> >You've already had an answer to the more general question.
> >
> >I hope this helps,
> >John
> >
> >--------------------------------
> >John Fox
> >Senator William McMaster
> >   Professor of Social Statistics
> >Department of Sociology
> >McMaster University
> >Hamilton, Ontario, Canada
> >http://socserv.mcmaster.ca/jfox
> >
> >
> >
> >
> > > -----Original Message-----
> > > From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
> > > project.org] On Behalf Of Kristi Shoemaker
> > > Sent: November-26-11 11:08 AM
> > > To: r-help at r-project.org
> > > Subject: [R] SPSS -> R
> > >
> > > I'm an SPSS user trying to make the transition to R.
> > >
> > > Can someone help me translate the following SPSS code into R?:
> > >
> > >
> > > GLM Total_tp1 Total_tp2 WITH Age Sex
> > >   /WSFACTOR=Time 2 Repeated
> > >   /METHOD=SSTYPE(3)
> > >   /CRITERIA=ALPHA(.05)
> > >   /WSDESIGN= Time
> > >   /DESIGN= Age Sex Age*Sex.
> > >
> > > Also. can anyone recommend any resources to help SPSS users learn
> to
> > > things in R?
> > >
> > > Thanks,
> > > -kristi
> > >
> > >     [[alternative HTML version deleted]]
> >         [[alternative HTML version deleted]]
> 
> Michael Dewey
> info at aghmed.fsnet.co.uk
> http://www.aghmed.fsnet.co.uk/home.html
> 
> ______________________________________________
> 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.



More information about the R-help mailing list