[R] MANOVA post hoc testing

John Fox jfox at mcmaster.ca
Fri Apr 4 13:55:17 CEST 2014

Dear Nicholas,

On Fri, 4 Apr 2014 04:59:09 +0000
 <Nicholas.Surawski at csiro.au> wrote:
> Greetings,
> I'm interested in performing some post hoc tests after conducting a multivariate analysis of covariance (MANCOVA) which I performed using the Anova function in the car package. The covariate did not end up being statistically significant, but the single factor's effect on the multivariate response was statistically significant.  From here, I would like to use the linearHypothesis function in the car package to test for where significant differences are occurring.
> Codewise, to fit the MANCOVA i've used:
> Model_1 <- Anova(lm((Y)~X+covariate))
> I've then tried to perform multiple comparisons using:
> linearHypothesis(Model_1,matchCoefs(Model_1,"X"),white.adjust=TRUE)

The argument white.adjust=TRUE isn't available for a multivariate linear model; see ?linearHypothesis. Also, the command (without the white.adjust argument) will give you the same test as Anova() did.


John Fox, Professor
McMaster University
Hamilton, Ontario, Canada
> After I run the 2nd bit of code I get the error message: " Error in vcov.default(model) :   there is no vcov() method for models of class Anova.mlm".
> Presumably, this means that I've stuffed up in trying to estimate the variance-covariance matrix for the Anova function.
> Any suggestions/help to resolve this problem would be greatly appreciated.
> Cheers,
> Nic Surawski
> 	[[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.

More information about the R-help mailing list