[R-sig-ME] Non-parametric 2x2 Mixed Factorial ANOVA in R?

Ben Bolker bbolker at gmail.com
Sun Nov 18 17:56:52 CET 2012


On 12-11-18 11:26 AM, Marianne Anke wrote:
> Thanks Ben and David for your help!
> 
> 
> I have done permutation testing and the results I got were actually
> similar compared to the traditional ANOVA (see below).
> 
> How should I phrase the result of this permutation test in an article?
> Would you mind making a suggestion?

  I would probably just present the results of the traditional ANOVA and
state in passing (maybe parenthetically) that the p-values derived from
a permutation-based analogue of the test were similar (and thus the
results are robust to non-normality of the data).  Don't forget to cite
as in citation("lmPerm") ...

> 
> 
> Thanks and best regards,
> 
> Marianne
> 
> 
> #####################################################################
> 
> Traditionalaov()ANOVA: 
> 
> 
> Error:Subject
> 
>           DfSumSqMeanSqFvaluePr(>F)
> 
> Group      1  17784  17784  0.972  0.337
> 
> Residuals 18 329459   18303               
> 
> 
> Error:Within
> 
>               Df Sum Sq Mean Sq F value Pr(>F)  
> 
> Session        1    831     831   0.408 0.5311  
> 
> Group:Session  1  10530  10530  5.1660.0355*
> 
> Residuals     18  36693    2038                 
> 
> ---
> 
> Signif.codes:  0‘***’0.001‘**’0.01‘*’0.05‘.’0.1‘’1 
> 
> 
> 
> Permutationaovp()ANOVA: 
> 
> [1]"Settings:  unique SS "
> 
> 
> Error:Subject
> 
> Component1:
> 
>           DfRSumSqRMeanSqIterPr(Prob)
> 
> Group      1    17784     17784  403    0.201
> 
> Residuals 18   329459     18303              
> 
> 
> 
> Error:Within
> 
> Component1:
> 
>               Df R Sum Sq R Mean Sq Iter Pr(Prob)  
> 
> Session        1      831     831.4  158  0.39241  
> 
> Group:Session  1    10530  10530.43934  0.02491*
> 
> Residuals     18    36693    2038.5                
> 
> ---
> 
> Signif.codes:  0‘***’0.001‘**’0.01‘*’0.05‘.’0.1‘’1 
> 
> #####################################################################
> 
> 
>> Date: Tue, 13 Nov 2012 07:50:05 +1000
>> From: David.Duffy at qimr.edu.au
>> To: r-sig-mixed-models at r-project.org
>> Subject: Re: [R-sig-ME] Non-parametric 2x2 Mixed Factorial ANOVA in R?
>>
>> On Tue, 13 Nov 2012, Ben Bolker wrote:
>>
>> > * you could try bootstrapping or permutation testing (being careful
>> > to respect the grouping structure of the data
>>
>> lmPerm package may be useful for this
>>
>> _______________________________________________
>> R-sig-mixed-models at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models



More information about the R-sig-mixed-models mailing list