WRS2 package. why do I only get one gender effect?

Hi there,

I hava a Bobodoll Experiment with Agression and I am looking for gender and Group differences. I failed my assumption testing and did a trimmed means ANOVA as seen below

Ge= Gender (factor)
gr = Group (factor)
physical aggression (int.)

t2way(phy_agg ~ Ge* gr, data = numbers)
test <- mcp2atm(phy_agg ~ Ge* gr, data = numbers)

Has anyone calculated this before? I am confused with the psihat for the mcp2atm post hoc test as I don't know how to do the symbol in r Markdown.

Another problem is that I only get one Ge1 when I call "test" and for the contrast, it's a matrix with only 1;0;-1. Does anyone know a paper order page explaining how to interpret these results?
my stat prof wanted to look over it but didn't have time yet and I was hoping some of you might have calculated this function before and could help me.

      psihat ci.lower     ci.upper    p-value

Ge1	-2.93838	-5.36048	   -0.51629	0.01818
gr1	-0.48384	-3.44648	    2.4788  0.69131
gr2	  -7.44444	  -9.51897	   -5.36991  0.00000
gr3	   -6.96061	    -9.1867     -4.73447 0.00000
Ge1:gr1	-0.95051	-3.91315	2.01214	0.43614
Ge1:gr2	-0.44444	-2.51897	1.63009	0.59486
Ge1:gr3	 0.50606	-1.72007	2.73220	0.57214
7 rows

Thank you!

EDIT: I have found the robust test from: Robust statistical methods in R using the WRS2 package, by Patrick Maier and Rand Wilcox

I am not familiar with this particular function, but I can give some feedback on how R produces stats outputs. Generally, if you have factor data (categories), one category is the "basis" against which the other categories are compared. If you haven't specified the order of your factors, R orders them alphabetically. So e.g. if you had the names female and male, female would be the "base" and the output will only show results or coefficients for male. So in your case, I suspect you have another factor in column Ge, which would be the "base", against which you should compare the results from Ge1.
So this is not a programming issue necessarily: you need to know/understand what kind of model you are building. You can find more information on the package you are using here: Robust statistical methods in R using the WRS2 package - Behavior Research Methods ( Mair, P., Wilcox, R. Robust statistical methods in R using the WRS2 package. Behav Res 52 , 464–488 (2020). Robust statistical methods in R using the WRS2 package - Behavior Research Methods). The formula for psihat is equation 11 in the article.

1 Like

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.